Datasets:

MaartenGr

/

arxiv_nlp

like 2

On the use of Silver Standard Data for Zero-shot Classification Tasks in Information Extraction

The superior performance of supervised classification methods in the information extraction (IE) area heavily relies on a large amount of gold standard data. Recent zero-shot classification methods converted the task to other NLP tasks (e.g., textual entailment) and used off-the-shelf models of these NLP tasks to directly perform inference on the test data without using a large amount of IE annotation data. A potentially valuable by-product of these methods is the large-scale silver standard data, i.e., pseudo-labeled data by the off-the-shelf models of other NLP tasks. However, there is no further investigation into the use of these data. In this paper, we propose a new framework, Clean-LaVe, which aims to utilize silver standard data to enhance the zero-shot performance. Clean-LaVe includes four phases: (1) Obtaining silver data; (2) Identifying relatively clean data from silver data; (3) Finetuning the off-the-shelf model using clean data; (4) Inference on the test data. The experimental results show that Clean-LaVe can outperform the baseline by 5% and 6% on TACRED and Wiki80 dataset in the zero-shot relation classification task, and by 3%-7% on Smile (Korean and Polish) in the zero-shot cross-lingual relation classification task, and by 8% on ACE05-E+ in the zero-shot event argument classification task. The code is share in https://github.com/wjw136/Clean_LaVe.git.

Computation and Language

Editing Factual Knowledge and Explanatory Ability of Medical Large Language Models

Model editing aims to precisely modify the behaviours of large language models (LLMs) on specific knowledge while keeping irrelevant knowledge unchanged. It has been proven effective in resolving hallucination and out-of-date issues in LLMs. As a result, it can boost the application of LLMs in many critical domains (e.g., medical domain), where the hallucination is not tolerable. In this paper, we propose two model editing studies and validate them in the medical domain: (1) directly editing the factual medical knowledge and (2) editing the explanations to facts. Meanwhile, we observed that current model editing methods struggle with the specialization and complexity of medical knowledge. Therefore, we propose MedLaSA, a novel Layer-wise Scalable Adapter strategy for medical model editing. It employs causal tracing to identify the precise location of knowledge in neurons and then introduces scalable adapters into the dense layers of LLMs. These adapters are assigned scaling values based on the corresponding specific knowledge. To evaluate the editing impact, we build two benchmark datasets and introduce a series of challenging and comprehensive metrics. Extensive experiments on medical LLMs demonstrate the editing efficiency of MedLaSA, without affecting irrelevant knowledge that is not edited.

Computation and Language

Assessing the Efficacy of Grammar Error Correction: A Human Evaluation Approach in the Japanese Context

In this study, we evaluated the performance of the state-of-the-art sequence tagging grammar error detection and correction model (SeqTagger) using Japanese university students' writing samples. With an automatic annotation toolkit, ERRANT, we first evaluated SeqTagger's performance on error correction with human expert correction as the benchmark. Then a human-annotated approach was adopted to evaluate Seqtagger's performance in error detection using a subset of the writing dataset. Results indicated a precision of 63.66% and a recall of 20.19% for error correction in the full dataset. For the subset, after manual exclusion of irrelevant errors such as semantic and mechanical ones, the model shows an adjusted precision of 97.98% and an adjusted recall of 42.98% for error detection, indicating the model's high accuracy but also its conservativeness. Thematic analysis on errors undetected by the model revealed that determiners and articles, especially the latter, were predominant. Specifically, in terms of context-independent errors, the model occasionally overlooked basic ones and faced challenges with overly erroneous or complex structures. Meanwhile, context-dependent errors, notably those related to tense and noun number, as well as those possibly influenced by the students' first language (L1), remained particularly challenging.

Computation and Language

Small But Funny: A Feedback-Driven Approach to Humor Distillation

The emergence of Large Language Models (LLMs) has brought to light promising language generation capabilities, particularly in performing tasks like complex reasoning and creative writing. Consequently, distillation through imitation of teacher responses has emerged as a popular technique to transfer knowledge from LLMs to more accessible, Small Language Models (SLMs). While this works well for simpler tasks, there is a substantial performance gap on tasks requiring intricate language comprehension and creativity, such as humor generation. We hypothesize that this gap may stem from the fact that creative tasks might be hard to learn by imitation alone and explore whether an approach, involving supplementary guidance from the teacher, could yield higher performance. To address this, we study the effect of assigning a dual role to the LLM - as a "teacher" generating data, as well as a "critic" evaluating the student's performance. Our experiments on humor generation reveal that the incorporation of feedback significantly narrows the performance gap between SLMs and their larger counterparts compared to merely relying on imitation. As a result, our research highlights the potential of using feedback as an additional dimension to data when transferring complex language abilities via distillation.

Computation and Language

Exploring Multilingual Human Value Concepts in Large Language Models: Is Value Alignment Consistent, Transferable and Controllable across Languages?

Prior research in representation engineering has revealed that LLMs encode concepts within their representation spaces, predominantly centered around English. In this study, we extend this philosophy to a multilingual scenario, delving into multilingual human value concepts in LLMs. Through our comprehensive exploration covering 7 types of human values, 16 languages and 3 LLM series with distinct multilinguality, we empirically substantiate the existence of multilingual human values in LLMs. Further cross-lingual analysis on these concepts discloses 3 traits arising from language resource disparities: cross-lingual inconsistency, distorted linguistic relationships, and unidirectional cross-lingual transfer between high- and low-resource languages, all in terms of human value concepts. Additionally, we validate the feasibility of cross-lingual control over value alignment capabilities of LLMs, leveraging the dominant language as a source language. Drawing from our findings on multilingual value alignment, we prudently provide suggestions on the composition of multilingual data for LLMs pre-training: including a limited number of dominant languages for cross-lingual alignment transfer while avoiding their excessive prevalence, and keeping a balanced distribution of non-dominant languages. We aspire that our findings would contribute to enhancing the safety and utility of multilingual AI.

Computation and Language

Saving the legacy of Hero Ibash: Evaluating Four Language Models for Aminoacian

This study assesses four cutting-edge language models in the underexplored Aminoacian language. Through evaluation, it scrutinizes their adaptability, effectiveness, and limitations in text generation, semantic coherence, and contextual understanding. Uncovering insights into these models' performance in a low-resourced language, this research pioneers pathways to bridge linguistic gaps. By offering benchmarks and understanding challenges, it lays groundwork for future advancements in natural language processing, aiming to elevate the applicability of language models in similar linguistic landscapes, marking a significant step toward inclusivity and progress in language technology.

Computation and Language

Cause and Effect: Can Large Language Models Truly Understand Causality?

With the rise of Large Language Models(LLMs), it has become crucial to understand their capabilities and limitations in deciphering and explaining the complex web of causal relationships that language entails. Current methods use either explicit or implicit causal reasoning, yet there is a strong need for a unified approach combining both to tackle a wide array of causal relationships more effectively. This research proposes a novel architecture called Context Aware Reasoning Enhancement with Counterfactual Analysis(CARE CA) framework to enhance causal reasoning and explainability. The proposed framework incorporates an explicit causal detection module with ConceptNet and counterfactual statements, as well as implicit causal detection through LLMs. Our framework goes one step further with a layer of counterfactual explanations to accentuate LLMs understanding of causality. The knowledge from ConceptNet enhances the performance of multiple causal reasoning tasks such as causal discovery, causal identification and counterfactual reasoning. The counterfactual sentences add explicit knowledge of the not caused by scenarios. By combining these powerful modules, our model aims to provide a deeper understanding of causal relationships, enabling enhanced interpretability. Evaluation of benchmark datasets shows improved performance across all metrics, such as accuracy, precision, recall, and F1 scores. We also introduce CausalNet, a new dataset accompanied by our code, to facilitate further research in this domain.

Computation and Language

Learning Intrinsic Dimension via Information Bottleneck for Explainable Aspect-based Sentiment Analysis

Gradient-based explanation methods are increasingly used to interpret neural models in natural language processing (NLP) due to their high fidelity. Such methods determine word-level importance using dimension-level gradient values through a norm function, often presuming equal significance for all gradient dimensions. However, in the context of Aspect-based Sentiment Analysis (ABSA), our preliminary research suggests that only specific dimensions are pertinent. To address this, we propose the Information Bottleneck-based Gradient (\texttt{IBG}) explanation framework for ABSA. This framework leverages an information bottleneck to refine word embeddings into a concise intrinsic dimension, maintaining essential features and omitting unrelated information. Comprehensive tests show that our \texttt{IBG} approach considerably improves both the models' performance and interpretability by identifying sentiment-aware features.

Computation and Language

Unsupervised Information Refinement Training of Large Language Models for Retrieval-Augmented Generation

Retrieval-augmented generation (RAG) enhances large language models (LLMs) by incorporating additional information from retrieval. However, studies have shown that LLMs still face challenges in effectively using the retrieved information, even ignoring it or being misled by it. The key reason is that the training of LLMs does not clearly make LLMs learn how to utilize input retrieved texts with varied quality. In this paper, we propose a novel perspective that considers the role of LLMs in RAG as ``Information Refiner'', which means that regardless of correctness, completeness, or usefulness of retrieved texts, LLMs can consistently integrate knowledge within the retrieved texts and model parameters to generate the texts that are more concise, accurate, and complete than the retrieved texts. To this end, we propose an information refinement training method named InFO-RAG that optimizes LLMs for RAG in an unsupervised manner. InFO-RAG is low-cost and general across various tasks. Extensive experiments on zero-shot prediction of 11 datasets in diverse tasks including Question Answering, Slot-Filling, Language Modeling, Dialogue, and Code Generation show that InFO-RAG improves the performance of LLaMA2 by an average of 9.39\% relative points. InFO-RAG also shows advantages in in-context learning and robustness of RAG.

Computation and Language

Cutting Off the Head Ends the Conflict: A Mechanism for Interpreting and Mitigating Knowledge Conflicts in Language Models

Recently, retrieval augmentation and tool augmentation have demonstrated a remarkable capability to expand the internal memory boundaries of language models (LMs) by providing external context. However, internal memory and external context inevitably clash, leading to knowledge conflicts within LMs. In this paper, we aim to interpret the mechanism of knowledge conflicts through the lens of information flow, and then mitigate conflicts by precise interventions at the pivotal point. We find there are some attention heads with opposite effects in the later layers, where memory heads can recall knowledge from internal memory, and context heads can retrieve knowledge from external context. Moreover, we reveal that the pivotal point at which knowledge conflicts emerge in LMs is the integration of inconsistent information flows by memory heads and context heads. Inspired by the insights, we propose a novel method called Pruning Head via PatH PatcHing (PH3), which can efficiently mitigate knowledge conflicts by pruning conflicting attention heads without updating model parameters. PH3 can flexibly control eight LMs to use internal memory ($\uparrow$ 44.0%) or external context ($\uparrow$ 38.5%). Moreover, PH3 can also improve the performance of LMs on open-domain QA tasks. We also conduct extensive experiments to demonstrate the cross-model, cross-relation, and cross-format generalization of our method.

Computation and Language

Evaluating Quantized Large Language Models

Post-training quantization (PTQ) has emerged as a promising technique to reduce the cost of large language models (LLMs). Specifically, PTQ can effectively mitigate memory consumption and reduce computational overhead in LLMs. To meet the requirements of both high efficiency and performance across diverse scenarios, a comprehensive evaluation of quantized LLMs is essential to guide the selection of quantization methods. This paper presents a thorough evaluation of these factors by evaluating the effect of PTQ on Weight, Activation, and KV Cache on 11 model families, including OPT, LLaMA2, Falcon, Bloomz, Mistral, ChatGLM, Vicuna, LongChat, StableLM, Gemma, and Mamba, with parameters ranging from 125M to 180B. The evaluation encompasses five types of tasks: basic NLP, emergent ability, trustworthiness, dialogue, and long-context tasks. Moreover, we also evaluate the state-of-the-art (SOTA) quantization methods to demonstrate their applicability. Based on the extensive experiments, we systematically summarize the effect of quantization, provide recommendations to apply quantization techniques, and point out future directions.

Computation and Language

MIKO: Multimodal Intention Knowledge Distillation from Large Language Models for Social-Media Commonsense Discovery

Social media has become a ubiquitous tool for connecting with others, staying updated with news, expressing opinions, and finding entertainment. However, understanding the intention behind social media posts remains challenging due to the implicitness of intentions in social media posts, the need for cross-modality understanding of both text and images, and the presence of noisy information such as hashtags, misspelled words, and complicated abbreviations. To address these challenges, we present MIKO, a Multimodal Intention Kowledge DistillatiOn framework that collaboratively leverages a Large Language Model (LLM) and a Multimodal Large Language Model (MLLM) to uncover users' intentions. Specifically, we use an MLLM to interpret the image and an LLM to extract key information from the text and finally instruct the LLM again to generate intentions. By applying MIKO to publicly available social media datasets, we construct an intention knowledge base featuring 1,372K intentions rooted in 137,287 posts. We conduct a two-stage annotation to verify the quality of the generated knowledge and benchmark the performance of widely used LLMs for intention generation. We further apply MIKO to a sarcasm detection dataset and distill a student model to demonstrate the downstream benefits of applying intention knowledge.

Computation and Language

Challenges in Pre-Training Graph Neural Networks for Context-Based Fake News Detection: An Evaluation of Current Strategies and Resource Limitations

Pre-training of neural networks has recently revolutionized the field of Natural Language Processing (NLP) and has before demonstrated its effectiveness in computer vision. At the same time, advances around the detection of fake news were mainly driven by the context-based paradigm, where different types of signals (e.g. from social media) form graph-like structures that hold contextual information apart from the news article to classify. We propose to merge these two developments by applying pre-training of Graph Neural Networks (GNNs) in the domain of context-based fake news detection. Our experiments provide an evaluation of different pre-training strategies for graph-based misinformation detection and demonstrate that transfer learning does currently not lead to significant improvements over training a model from scratch in the domain. We argue that a major current issue is the lack of suitable large-scale resources that can be used for pre-training.

Computation and Language

Clustering and Ranking: Diversity-preserved Instruction Selection through Expert-aligned Quality Estimation

With contributions from the open-source community, a vast amount of instruction tuning (IT) data has emerged. Given the significant resource allocation required by training and evaluating models, it is advantageous to have an efficient method for selecting high-quality IT data. However, existing methods for instruction data selection have limitations such as relying on fragile external APIs, being affected by biases in GPT models, or reducing the diversity of the selected instruction dataset. In this paper, we propose an industrial-friendly, expert-aligned and diversity-preserved instruction data selection method: Clustering and Ranking (CaR). CaR consists of two steps. The first step involves ranking instruction pairs using a scoring model that is well aligned with expert preferences (achieving an accuracy of 84.25%). The second step involves preserving dataset diversity through a clustering process.In our experiment, CaR selected a subset containing only 1.96% of Alpaca's IT data, yet the underlying AlpaCaR model trained on this subset outperforms Alpaca by an average of 32.1% in GPT-4 evaluations. Furthermore, our method utilizes small models (355M parameters) and requires only 11.2% of the monetary cost compared to existing methods, making it easily deployable in industrial scenarios.

Computation and Language

DANSK and DaCy 2.6.0: Domain Generalization of Danish Named Entity Recognition

Named entity recognition is one of the cornerstones of Danish NLP, essential for language technology applications within both industry and research. However, Danish NER is inhibited by a lack of available datasets. As a consequence, no current models are capable of fine-grained named entity recognition, nor have they been evaluated for potential generalizability issues across datasets and domains. To alleviate these limitations, this paper introduces: 1) DANSK: a named entity dataset providing for high-granularity tagging as well as within-domain evaluation of models across a diverse set of domains; 2) DaCy 2.6.0 that includes three generalizable models with fine-grained annotation; and 3) an evaluation of current state-of-the-art models' ability to generalize across domains. The evaluation of existing and new models revealed notable performance discrepancies across domains, which should be addressed within the field. Shortcomings of the annotation quality of the dataset and its impact on model training and evaluation are also discussed. Despite these limitations, we advocate for the use of the new dataset DANSK alongside further work on the generalizability within Danish NER.

Computation and Language

LLM Task Interference: An Initial Study on the Impact of Task-Switch in Conversational History

With the recent emergence of powerful instruction-tuned large language models (LLMs), various helpful conversational Artificial Intelligence (AI) systems have been deployed across many applications. When prompted by users, these AI systems successfully perform a wide range of tasks as part of a conversation. To provide some sort of memory and context, such approaches typically condition their output on the entire conversational history. Although this sensitivity to the conversational history can often lead to improved performance on subsequent tasks, we find that performance can in fact also be negatively impacted, if there is a task-switch. To the best of our knowledge, our work makes the first attempt to formalize the study of such vulnerabilities and interference of tasks in conversational LLMs caused by task-switches in the conversational history. Our experiments across 5 datasets with 15 task switches using popular LLMs reveal that many of the task-switches can lead to significant performance degradation.

Computation and Language

Improving Open-Ended Text Generation via Adaptive Decoding

Current language models decode text token by token according to probabilistic distribution, and determining the appropriate candidates for the next token is crucial to ensure generation quality. This study introduces adaptive decoding, a mechanism that empowers the language models to ascertain a sensible candidate set during the generation process dynamically. Specifically, we introduce an entropy-based metric called confidence and conceptualize determining the optimal candidate set as a confidence-increasing process. The rationality of including a token in the candidate set is assessed by leveraging the increment of confidence, enabling the model to determine the most suitable candidate set adaptively. The experimental results reveal that our method achieves higher MAUVE and diversity in story generation tasks and maintains certain coherence, underscoring its superiority over existing algorithms. The code is available at https://github.com/zwhong714/adaptive_decoding.

Computation and Language

CogBench: a large language model walks into a psychology lab

Large language models (LLMs) have significantly advanced the field of artificial intelligence. Yet, evaluating them comprehensively remains challenging. We argue that this is partly due to the predominant focus on performance metrics in most benchmarks. This paper introduces CogBench, a benchmark that includes ten behavioral metrics derived from seven cognitive psychology experiments. This novel approach offers a toolkit for phenotyping LLMs' behavior. We apply CogBench to 35 LLMs, yielding a rich and diverse dataset. We analyze this data using statistical multilevel modeling techniques, accounting for the nested dependencies among fine-tuned versions of specific LLMs. Our study highlights the crucial role of model size and reinforcement learning from human feedback (RLHF) in improving performance and aligning with human behavior. Interestingly, we find that open-source models are less risk-prone than proprietary models and that fine-tuning on code does not necessarily enhance LLMs' behavior. Finally, we explore the effects of prompt-engineering techniques. We discover that chain-of-thought prompting improves probabilistic reasoning, while take-a-step-back prompting fosters model-based behaviors.

Computation and Language

Learning or Self-aligning? Rethinking Instruction Fine-tuning

Instruction Fine-tuning~(IFT) is a critical phase in building large language models~(LLMs). Previous works mainly focus on the IFT's role in the transfer of behavioral norms and the learning of additional world knowledge. However, the understanding of the underlying mechanisms of IFT remains significantly limited. In this paper, we design a knowledge intervention framework to decouple the potential underlying factors of IFT, thereby enabling individual analysis of different factors. Surprisingly, our experiments reveal that attempting to learn additional world knowledge through IFT often struggles to yield positive impacts and can even lead to markedly negative effects. Further, we discover that maintaining internal knowledge consistency before and after IFT is a critical factor for achieving successful IFT. Our findings reveal the underlying mechanisms of IFT and provide robust support for some very recent and potential future works.

Computation and Language

Towards Generalist Prompting for Large Language Models by Mental Models

Large language models (LLMs) have demonstrated impressive performance on many tasks. However, to achieve optimal performance, specially designed prompting methods are still needed. These methods either rely on task-specific few-shot examples that require a certain level of domain knowledge, or are designed to be simple but only perform well on a few types of tasks. In this work, we attempt to introduce the concept of generalist prompting, which operates on the design principle of achieving optimal or near-optimal performance on a wide range of tasks while eliminating the need for manual selection and customization of prompts tailored to specific problems. Furthermore, we propose MeMo (Mental Models), an innovative prompting method that is simple-designed yet effectively fulfills the criteria of generalist prompting. MeMo distills the cores of various prompting methods into individual mental models and allows LLMs to autonomously select the most suitable mental models for the problem, achieving or being near to the state-of-the-art results on diverse tasks such as STEM, logical reasoning, and commonsense reasoning in zero-shot settings. We hope that the insights presented herein will stimulate further exploration of generalist prompting methods for LLMs.

Computation and Language

A BiRGAT Model for Multi-intent Spoken Language Understanding with Hierarchical Semantic Frames

Previous work on spoken language understanding (SLU) mainly focuses on single-intent settings, where each input utterance merely contains one user intent. This configuration significantly limits the surface form of user utterances and the capacity of output semantics. In this work, we first propose a Multi-Intent dataset which is collected from a realistic in-Vehicle dialogue System, called MIVS. The target semantic frame is organized in a 3-layer hierarchical structure to tackle the alignment and assignment problems in multi-intent cases. Accordingly, we devise a BiRGAT model to encode the hierarchy of ontology items, the backbone of which is a dual relational graph attention network. Coupled with the 3-way pointer-generator decoder, our method outperforms traditional sequence labeling and classification-based schemes by a large margin.

Computation and Language

Hierarchical Multimodal Pre-training for Visually Rich Webpage Understanding

The growing prevalence of visually rich documents, such as webpages and scanned/digital-born documents (images, PDFs, etc.), has led to increased interest in automatic document understanding and information extraction across academia and industry. Although various document modalities, including image, text, layout, and structure, facilitate human information retrieval, the interconnected nature of these modalities presents challenges for neural networks. In this paper, we introduce WebLM, a multimodal pre-training network designed to address the limitations of solely modeling text and structure modalities of HTML in webpages. Instead of processing document images as unified natural images, WebLM integrates the hierarchical structure of document images to enhance the understanding of markup-language-based documents. Additionally, we propose several pre-training tasks to model the interaction among text, structure, and image modalities effectively. Empirical results demonstrate that the pre-trained WebLM significantly surpasses previous state-of-the-art pre-trained models across several webpage understanding tasks. The pre-trained models and code are available at https://github.com/X-LANCE/weblm.

Computation and Language

Retrieval-based Full-length Wikipedia Generation for Emergent Events

In today's fast-paced world, the growing demand to quickly generate comprehensive and accurate Wikipedia documents for emerging events is both crucial and challenging. However, previous efforts in Wikipedia generation have often fallen short of meeting real-world requirements. Some approaches focus solely on generating segments of a complete Wikipedia document, while others overlook the importance of faithfulness in generation or fail to consider the influence of the pre-training corpus. In this paper, we simulate a real-world scenario where structured full-length Wikipedia documents are generated for emergent events using input retrieved from web sources. To ensure that Large Language Models (LLMs) are not trained on corpora related to recently occurred events, we select events that have taken place recently and introduce a new benchmark Wiki-GenBen, which consists of 309 events paired with their corresponding retrieved web pages for generating evidence. Additionally, we design a comprehensive set of systematic evaluation metrics and baseline methods, to evaluate the capability of LLMs in generating factual full-length Wikipedia documents. The data and code are open-sourced at WikiGenBench.

Computation and Language

A Survey on Neural Question Generation: Methods, Applications, and Prospects

In this survey, we present a detailed examination of the advancements in Neural Question Generation (NQG), a field leveraging neural network techniques to generate relevant questions from diverse inputs like knowledge bases, texts, and images. The survey begins with an overview of NQG's background, encompassing the task's problem formulation, prevalent benchmark datasets, established evaluation metrics, and notable applications. It then methodically classifies NQG approaches into three predominant categories: structured NQG, which utilizes organized data sources, unstructured NQG, focusing on more loosely structured inputs like texts or visual content, and hybrid NQG, drawing on diverse input modalities. This classification is followed by an in-depth analysis of the distinct neural network models tailored for each category, discussing their inherent strengths and potential limitations. The survey culminates with a forward-looking perspective on the trajectory of NQG, identifying emergent research trends and prospective developmental paths. Accompanying this survey is a curated collection of related research papers, datasets and codes, systematically organized on Github, providing an extensive reference for those delving into NQG.

Computation and Language

Rethinking the Bounds of LLM Reasoning: Are Multi-Agent Discussions the Key?

Recent progress in LLMs discussion suggests that multi-agent discussion improves the reasoning abilities of LLMs. In this work, we reevaluate this claim through systematic experiments, where we propose a novel group discussion framework to enrich the set of discussion mechanisms. Interestingly, our results show that a single-agent LLM with strong prompts can achieve almost the same performance as the best existing discussion approach on a wide range of reasoning tasks and backbone LLMs. We observe that the multi-agent discussion performs better than a single agent only when there is no demonstration in the prompt. Further study reveals the common interaction mechanisms of LLMs during the discussion.

Computation and Language

Towards Better Understanding of Contrastive Sentence Representation Learning: A Unified Paradigm for Gradient

Sentence Representation Learning (SRL) is a crucial task in Natural Language Processing (NLP), where contrastive Self-Supervised Learning (SSL) is currently a mainstream approach. However, the reasons behind its remarkable effectiveness remain unclear. Specifically, in other research fields, contrastive SSL shares similarities in both theory and practical performance with non-contrastive SSL (e.g., alignment & uniformity, Barlow Twins, and VICReg). However, in SRL, contrastive SSL outperforms non-contrastive SSL significantly. Therefore, two questions arise: First, what commonalities enable various contrastive losses to achieve superior performance in SRL? Second, how can we make non-contrastive SSL, which is similar to contrastive SSL but ineffective in SRL, effective? To address these questions, we start from the perspective of gradients and discover that four effective contrastive losses can be integrated into a unified paradigm, which depends on three components: the Gradient Dissipation, the Weight, and the Ratio. Then, we conduct an in-depth analysis of the roles these components play in optimization and experimentally demonstrate their significance for model performance. Finally, by adjusting these components, we enable non-contrastive SSL to achieve outstanding performance in SRL.

Computation and Language

Is Crowdsourcing Breaking Your Bank? Cost-Effective Fine-Tuning of Pre-trained Language Models with Proximal Policy Optimization

Wide usage of ChatGPT has highlighted the potential of reinforcement learning from human feedback. However, its training pipeline relies on manual ranking, a resource-intensive process. To reduce labor costs, we propose a self-supervised text ranking approach for applying Proximal-Policy-Optimization to fine-tune language models while eliminating the need for human annotators. Our method begins with probabilistic sampling to encourage a language model to generate diverse responses for each input. We then employ TextRank and ISODATA algorithms to rank and cluster these responses based on their semantics. Subsequently, we construct a reward model to learn the rank and optimize our generative policy. Our experimental results, conducted using two language models on three tasks, demonstrate that the models trained by our method considerably outperform baselines regarding BLEU, GLEU, and METEOR scores. Furthermore, our manual evaluation shows that our ranking results exhibit a remarkably high consistency with that of humans. This research significantly reduces training costs of proximal policy-guided models and demonstrates the potential for self-correction of language models.

Computation and Language

How to think step-by-step: A mechanistic understanding of chain-of-thought reasoning

Despite superior reasoning prowess demonstrated by Large Language Models (LLMs) with Chain-of-Thought (CoT) prompting, a lack of understanding prevails around the internal mechanisms of the models that facilitate CoT generation. This work investigates the neural sub-structures within LLMs that manifest CoT reasoning from a mechanistic point of view. From an analysis of LLaMA-2 7B applied to multistep reasoning over fictional ontologies, we demonstrate that LLMs deploy multiple parallel pathways of answer generation for step-by-step reasoning. These parallel pathways provide sequential answers from the input question context as well as the generated CoT. We observe a striking functional rift in the middle layers of the LLM. Token representations in the initial half remain strongly biased towards the pretraining prior, with the in-context taking over abruptly in the later half. This internal phase shift manifests in different functional components: attention heads that write the answer token predominantly appear in the later half, attention heads that move information along ontological relationships appear exclusively in the initial half, and so on. To the best of our knowledge, this is the first attempt towards mechanistic investigation of CoT reasoning in LLMs.

Computation and Language

Learning to Generate Instruction Tuning Datasets for Zero-Shot Task Adaptation

We introduce Bonito, an open-source model for conditional task generation: the task of converting unannotated text into task-specific training datasets for instruction tuning. Our goal is to enable zero-shot task adaptation of large language models on users' specialized, private data. We train Bonito on a new large-scale dataset with 1.65M examples created by remixing existing instruction tuning datasets into meta-templates. The meta-templates for a dataset produce training examples where the input is the unannotated text and the task attribute and the output consists of the instruction and the response. We use Bonito to generate synthetic tasks for seven datasets from specialized domains across three task types -- yes-no question answering, extractive question answering, and natural language inference -- and adapt language models. We show that Bonito significantly improves the average performance of pretrained and instruction tuned models over the de facto self supervised baseline. For example, adapting Mistral-Instruct-v2 and instruction tuned variants of Mistral and Llama2 with Bonito improves the strong zero-shot performance by 22.1 F1 points whereas the next word prediction objective undoes some of the benefits of instruction tuning and reduces the average performance by 0.8 F1 points. We conduct additional experiments with Bonito to understand the effects of the domain, the size of the training set, and the choice of alternative synthetic task generators. Overall, we show that learning with synthetic instruction tuning datasets is an effective way to adapt language models to new domains. The model, dataset, and code are available at https://github.com/BatsResearch/bonito.

Computation and Language

Focus on Your Question! Interpreting and Mitigating Toxic CoT Problems in Commonsense Reasoning

Large language models exhibit high-level commonsense reasoning abilities, especially with enhancement methods like Chain-of-Thought (CoT). However, we find these CoT-like methods lead to a considerable number of originally correct answers turning wrong, which we define as the Toxic CoT problem. To interpret and mitigate this problem, we first utilize attribution tracing and causal tracing methods to probe the internal working mechanism of the LLM during CoT reasoning. Through comparisons, we prove that the model exhibits information loss from the question over the shallow attention layers when generating rationales or answers. Based on the probing findings, we design a novel method called RIDERS (Residual decodIng and sERial-position Swap), which compensates for the information deficit in the model from both decoding and serial-position perspectives. Through extensive experiments on multiple commonsense reasoning benchmarks, we validate that this method not only significantly eliminates Toxic CoT problems (decreased by 23.6%), but also effectively improves the model's overall commonsense reasoning performance (increased by 5.5%).

Computation and Language

VerifiNER: Verification-augmented NER via Knowledge-grounded Reasoning with Large Language Models

Recent approaches in domain-specific named entity recognition (NER), such as biomedical NER, have shown remarkable advances. However, they still lack of faithfulness, producing erroneous predictions. We assume that knowledge of entities can be useful in verifying the correctness of the predictions. Despite the usefulness of knowledge, resolving such errors with knowledge is nontrivial, since the knowledge itself does not directly indicate the ground-truth label. To this end, we propose VerifiNER, a post-hoc verification framework that identifies errors from existing NER methods using knowledge and revises them into more faithful predictions. Our framework leverages the reasoning abilities of large language models to adequately ground on knowledge and the contextual information in the verification process. We validate effectiveness of VerifiNER through extensive experiments on biomedical datasets. The results suggest that VerifiNER can successfully verify errors from existing models as a model-agnostic approach. Further analyses on out-of-domain and low-resource settings show the usefulness of VerifiNER on real-world applications.

Computation and Language

Tokenization Is More Than Compression

Tokenization is a foundational step in Natural Language Processing (NLP) tasks, bridging raw text and language models. Existing tokenization approaches like Byte-Pair Encoding (BPE) originate from the field of data compression, and it has been suggested that the effectiveness of BPE stems from its ability to condense text into a relatively small number of tokens. We test the hypothesis that fewer tokens lead to better downstream performance by introducing PathPiece, a new tokenizer that segments a document's text into the minimum number of tokens for a given vocabulary. Through extensive experimentation we find this hypothesis not to be the case, casting doubt on the understanding of the reasons for effective tokenization. To examine which other factors play a role, we evaluate design decisions across all three phases of tokenization: pre-tokenization, vocabulary construction, and segmentation, offering new insights into the design of effective tokenizers. Specifically, we illustrate the importance of pre-tokenization and the benefits of using BPE to initialize vocabulary construction. We train 64 language models with varying tokenization, ranging in size from 350M to 2.4B parameters, all of which are made publicly available.

Computation and Language

The First Place Solution of WSDM Cup 2024: Leveraging Large Language Models for Conversational Multi-Doc QA

Conversational multi-doc question answering aims to answer specific questions based on the retrieved documents as well as the contextual conversations. In this paper, we introduce our winning approach for the "Conversational Multi-Doc QA" challenge in WSDM Cup 2024, which exploits the superior natural language understanding and generation capability of Large Language Models (LLMs). We first adapt LLMs to the task, then devise a hybrid training strategy to make the most of in-domain unlabeled data. Moreover, an advanced text embedding model is adopted to filter out potentially irrelevant documents and several approaches are designed and compared for the model ensemble. Equipped with all these techniques, our solution finally ranked 1st place in WSDM Cup 2024, surpassing its rivals to a large extent. The source codes have been released at https://github.com/zhangzhao219/WSDM-Cup-2024.

Computation and Language

Decomposed Prompting: Unveiling Multilingual Linguistic Structure Knowledge in English-Centric Large Language Models

Despite the predominance of English in their training data, English-centric Large Language Models (LLMs) like GPT-3 and LLaMA display a remarkable ability to perform multilingual tasks, raising questions about the depth and nature of their cross-lingual capabilities. This paper introduces the decomposed prompting approach to probe the linguistic structure understanding of these LLMs in sequence labeling tasks. Diverging from the single text-to-text prompt, our method generates for each token of the input sentence an individual prompt which asks for its linguistic label. We assess our method on the Universal Dependencies part-of-speech tagging dataset for 38 languages, utilizing both English-centric and multilingual LLMs. Our findings show that decomposed prompting surpasses the iterative prompting baseline in efficacy and efficiency under zero- and few-shot settings. Further analysis reveals the influence of evaluation methods and the use of instructions in prompts. Our multilingual investigation shows that English-centric language models perform better on average than multilingual models. Our study offers insights into the multilingual transferability of English-centric LLMs, contributing to the understanding of their multilingual linguistic knowledge.

Computation and Language

Can GPT Improve the State of Prior Authorization via Guideline Based Automated Question Answering?

Health insurance companies have a defined process called prior authorization (PA) which is a health plan cost-control process that requires doctors and other healthcare professionals to get clearance in advance from a health plan before performing a particular procedure on a patient in order to be eligible for payment coverage. For health insurance companies, approving PA requests for patients in the medical domain is a time-consuming and challenging task. One of those key challenges is validating if a request matches up to certain criteria such as age, gender, etc. In this work, we evaluate whether GPT can validate numerous key factors, in turn helping health plans reach a decision drastically faster. We frame it as a question answering task, prompting GPT to answer a question from patient electronic health record. We experiment with different conventional prompting techniques as well as introduce our own novel prompting technique. Moreover, we report qualitative assessment by humans on the natural language generation outputs from our approach. Results show that our method achieves superior performance with the mean weighted F1 score of 0.61 as compared to its standard counterparts.

Computation and Language

Emotion Classification in Low and Moderate Resource Languages

It is important to be able to analyze the emotional state of people around the globe. There are 7100+ active languages spoken around the world and building emotion classification for each language is labor intensive. Particularly for low-resource and endangered languages, building emotion classification can be quite challenging. We present a cross-lingual emotion classifier, where we train an emotion classifier with resource-rich languages (i.e. \textit{English} in our work) and transfer the learning to low and moderate resource languages. We compare and contrast two approaches of transfer learning from a high-resource language to a low or moderate-resource language. One approach projects the annotation from a high-resource language to low and moderate-resource language in parallel corpora and the other one uses direct transfer from high-resource language to the other languages. We show the efficacy of our approaches on 6 languages: Farsi, Arabic, Spanish, Ilocano, Odia, and Azerbaijani. Our results indicate that our approaches outperform random baselines and transfer emotions across languages successfully. For all languages, the direct cross-lingual transfer of emotion yields better results. We also create annotated emotion-labeled resources for four languages: Farsi, Azerbaijani, Ilocano and Odia.

Computation and Language

Leveraging Diverse Modeling Contexts with Collaborating Learning for Neural Machine Translation

Autoregressive (AR) and Non-autoregressive (NAR) models are two types of generative models for Neural Machine Translation (NMT). AR models predict tokens in a word-by-word manner and can effectively capture the distribution of real translations. NAR models predict tokens by extracting bidirectional contextual information which can improve the inference speed but they suffer from performance degradation. Previous works utilized AR models to enhance NAR models by reducing the training data's complexity or incorporating the global information into AR models by virtue of NAR models. However, those investigated methods only take advantage of the contextual information of a single type of model while neglecting the diversity in the contextual information that can be provided by different types of models. In this paper, we propose a novel generic collaborative learning method, DCMCL, where AR and NAR models are treated as collaborators instead of teachers and students. To hierarchically leverage the bilateral contextual information, token-level mutual learning and sequence-level contrastive learning are adopted between AR and NAR models. Extensive experiments on four widely used benchmarks show that the proposed DCMCL method can simultaneously improve both AR and NAR models with up to 1.38 and 2.98 BLEU scores respectively, and can also outperform the current best-unified model with up to 0.97 BLEU scores for both AR and NAR decoding.

Computation and Language

Beyond Natural Language: LLMs Leveraging Alternative Formats for Enhanced Reasoning and Communication

Natural language (NL) has long been the predominant format for human cognition and communication, and by extension, has been similarly pivotal in the development and application of Large Language Models (LLMs). Yet, besides NL, LLMs have seen various non-NL formats during pre-training, such as code and logical expression. NL's status as the optimal format for LLMs, particularly in single-LLM reasoning and multi-agent communication, has not been thoroughly examined. In this work, we challenge the default use of NL by exploring the utility of non-NL formats in these contexts. We show that allowing LLMs to autonomously select the most suitable format before reasoning or communicating leads to a 3.3 to 5.7\% improvement in reasoning efficiency for different LLMs, and up to a 72.7\% reduction in token usage in multi-agent communication, all while maintaining communicative effectiveness. Our comprehensive analysis further reveals that LLMs can devise a format from limited task instructions and that the devised format is effectively transferable across different LLMs. Intriguingly, the structured communication format decided by LLMs exhibits notable parallels with established agent communication languages, suggesting a natural evolution towards efficient, structured communication in agent communication. Our code is released at \url{https://github.com/thunlp/AutoForm}.

Computation and Language

HOP to the Next Tasks and Domains for Continual Learning in NLP

Continual Learning (CL) aims to learn a sequence of problems (i.e., tasks and domains) by transferring knowledge acquired on previous problems, whilst avoiding forgetting of past ones. Different from previous approaches which focused on CL for one NLP task or domain in a specific use-case, in this paper, we address a more general CL setting to learn from a sequence of problems in a unique framework. Our method, HOP, permits to hop across tasks and domains by addressing the CL problem along three directions: (i) we employ a set of adapters to generalize a large pre-trained model to unseen problems, (ii) we compute high-order moments over the distribution of embedded representations to distinguish independent and correlated statistics across different tasks and domains, (iii) we process this enriched information with auxiliary heads specialized for each end problem. Extensive experimental campaign on 4 NLP applications, 5 benchmarks and 2 CL setups demonstrates the effectiveness of our HOP.

Computation and Language

Meta-Task Prompting Elicits Embedding from Large Language Models

In this work, we introduce a new unsupervised embedding method, Meta-Task Prompting with Explicit One-Word Limitation (MetaEOL), for generating high-quality sentence embeddings from Large Language Models (LLMs) without the need for model fine-tuning or task-specific engineering. Leveraging meta-task prompting, MetaEOL guides LLMs to produce embeddings through a series of carefully designed prompts that address multiple representational aspects. Our comprehensive experiments demonstrate that embeddings averaged from various meta-tasks yield competitive performance on Semantic Textual Similarity (STS) benchmarks and excel in downstream tasks, surpassing contrastive-trained models. Our findings suggest a new scaling law for embedding generation, offering a versatile, resource-efficient approach for embedding extraction across diverse sentence-centric scenarios.

Computation and Language

NewsQs: Multi-Source Question Generation for the Inquiring Mind

We present NewsQs (news-cues), a dataset that provides question-answer pairs for multiple news documents. To create NewsQs, we augment a traditional multi-document summarization dataset with questions automatically generated by a T5-Large model fine-tuned on FAQ-style news articles from the News On the Web corpus. We show that fine-tuning a model with control codes produces questions that are judged acceptable more often than the same model without them as measured through human evaluation. We use a QNLI model with high correlation with human annotations to filter our data. We release our final dataset of high-quality questions, answers, and document clusters as a resource for future work in query-based multi-document summarization.

Computation and Language

Few-Shot Fairness: Unveiling LLM's Potential for Fairness-Aware Classification

Employing Large Language Models (LLM) in various downstream applications such as classification is crucial, especially for smaller companies lacking the expertise and resources required for fine-tuning a model. Fairness in LLMs helps ensure inclusivity, equal representation based on factors such as race, gender and promotes responsible AI deployment. As the use of LLMs has become increasingly prevalent, it is essential to assess whether LLMs can generate fair outcomes when subjected to considerations of fairness. In this study, we introduce a framework outlining fairness regulations aligned with various fairness definitions, with each definition being modulated by varying degrees of abstraction. We explore the configuration for in-context learning and the procedure for selecting in-context demonstrations using RAG, while incorporating fairness rules into the process. Experiments conducted with different LLMs indicate that GPT-4 delivers superior results in terms of both accuracy and fairness compared to other models. This work is one of the early attempts to achieve fairness in prediction tasks by utilizing LLMs through in-context learning.

Computation and Language

Large Language Models and Games: A Survey and Roadmap

Recent years have seen an explosive increase in research on large language models (LLMs), and accompanying public engagement on the topic. While starting as a niche area within natural language processing, LLMs have shown remarkable potential across a broad range of applications and domains, including games. This paper surveys the current state of the art across the various applications of LLMs in and for games, and identifies the different roles LLMs can take within a game. Importantly, we discuss underexplored areas and promising directions for future uses of LLMs in games and we reconcile the potential and limitations of LLMs within the games domain. As the first comprehensive survey and roadmap at the intersection of LLMs and games, we are hopeful that this paper will serve as the basis for groundbreaking research and innovation in this exciting new field.

Computation and Language

FOFO: A Benchmark to Evaluate LLMs' Format-Following Capability

This paper presents FoFo, a pioneering benchmark for evaluating large language models' (LLMs) ability to follow complex, domain-specific formats, a crucial yet underexamined capability for their application as AI agents. Despite LLMs' advancements, existing benchmarks fail to assess their format-following proficiency adequately. FoFo fills this gap with a diverse range of real-world formats and instructions, developed through an AI-Human collaborative method. Our evaluation across both open-source (e.g., Llama 2, WizardLM) and closed-source (e.g., GPT-4, PALM2, Gemini) LLMs highlights three key findings: open-source models significantly lag behind closed-source ones in format adherence; LLMs' format-following performance is independent of their content generation quality; and LLMs' format proficiency varies across different domains. These insights suggest the need for specialized tuning for format-following skills and highlight FoFo's role in guiding the selection of domain-specific AI agents. FoFo is released here at https://github.com/SalesforceAIResearch/FoFo.

Computation and Language

Simple linear attention language models balance the recall-throughput tradeoff

Recent work has shown that attention-based language models excel at recall, the ability to ground generations in tokens previously seen in context. However, the efficiency of attention-based models is bottle-necked during inference by the KV-cache's aggressive memory consumption. In this work, we explore whether we can improve language model efficiency (e.g. by reducing memory consumption) without compromising on recall. By applying experiments and theory to a broad set of architectures, we identify a key tradeoff between a model's state size and recall ability. We show that efficient alternatives to attention (e.g. H3, Mamba, RWKV) maintain a fixed-size recurrent state, but struggle at recall. We propose BASED a simple architecture combining linear and sliding window attention. By varying BASED window size and linear attention feature dimension, we can dial the state size and traverse the pareto frontier of the recall-memory tradeoff curve, recovering the full quality of attention on one end and the small state size of attention-alternatives on the other. We train language models up to 1.3b parameters and show that BASED matches the strongest sub-quadratic models (e.g. Mamba) in perplexity and outperforms them on real-world recall-intensive tasks by 6.22 accuracy points. Implementations of linear attention are often less efficient than optimized standard attention implementations. To make BASED competitive, we develop IO-aware algorithms that enable 24x higher throughput on language generation than FlashAttention-2, when generating 1024 tokens using 1.3b parameter models. Code for this work is provided at: https://github.com/HazyResearch/based.

Computation and Language

RORA: Robust Free-Text Rationale Evaluation

Free-text rationales play a pivotal role in explainable NLP, bridging the knowledge and reasoning gaps behind a model's decision-making. However, due to the diversity of potential reasoning paths and a corresponding lack of definitive ground truth, their evaluation remains a challenge. Existing evaluation metrics rely on the degree to which a rationale supports a target label, but we find these fall short in evaluating rationales that inadvertently leak the labels. To address this problem, we propose RORA, a Robust free-text Rationale evaluation against label leakage. RORA quantifies the new information supplied by a rationale to justify the label. This is achieved by assessing the conditional V-information \citep{hewitt-etal-2021-conditional} with a predictive family robust against leaky features that can be exploited by a small model. RORA consistently outperforms existing approaches in evaluating human-written, synthetic, or model-generated rationales, particularly demonstrating robustness against label leakage. We also show that RORA aligns well with human judgment, providing a more reliable and accurate measurement across diverse free-text rationales.

Computation and Language

Learning to Compress Prompt in Natural Language Formats

Large language models (LLMs) are great at processing multiple natural language processing tasks, but their abilities are constrained by inferior performance with long context, slow inference speed, and the high cost of computing the results. Deploying LLMs with precise and informative context helps users process large-scale datasets more effectively and cost-efficiently. Existing works rely on compressing long prompt contexts into soft prompts. However, soft prompt compression encounters limitations in transferability across different LLMs, especially API-based LLMs. To this end, this work aims to compress lengthy prompts in the form of natural language with LLM transferability. This poses two challenges: (i) Natural Language (NL) prompts are incompatible with back-propagation, and (ii) NL prompts lack flexibility in imposing length constraints. In this work, we propose a Natural Language Prompt Encapsulation (Nano-Capsulator) framework compressing original prompts into NL formatted Capsule Prompt while maintaining the prompt utility and transferability. Specifically, to tackle the first challenge, the Nano-Capsulator is optimized by a reward function that interacts with the proposed semantics preserving loss. To address the second question, the Nano-Capsulator is optimized by a reward function featuring length constraints. Experimental results demonstrate that the Capsule Prompt can reduce 81.4% of the original length, decrease inference latency up to 4.5x, and save 80.1% of budget overheads while providing transferability across diverse LLMs and different datasets.

Computation and Language

Fine-Tuned Machine Translation Metrics Struggle in Unseen Domains

We introduce a new, extensive multidimensional quality metrics (MQM) annotated dataset covering 11 language pairs in the biomedical domain. We use this dataset to investigate whether machine translation (MT) metrics which are fine-tuned on human-generated MT quality judgements are robust to domain shifts between training and inference. We find that fine-tuned metrics exhibit a substantial performance drop in the unseen domain scenario relative to metrics that rely on the surface form, as well as pre-trained metrics which are not fine-tuned on MT quality judgments.

Computation and Language

How Much Annotation is Needed to Compare Summarization Models?

Modern instruction-tuned models have become highly capable in text generation tasks such as summarization, and are expected to be released at a steady pace. In practice one may now wish to choose confidently, but with minimal effort, the best performing summarization model when applied to a new domain or purpose. In this work, we empirically investigate the test sample size necessary to select a preferred model in the context of news summarization. Empirical results reveal that comparative evaluation converges quickly for both automatic and human evaluation, with clear preferences for a system emerging from under 100 examples. The human preference data allows us to quantify how well automatic scores can reproduce preference rankings across a variety of downstream summarization tasks. We find that, while automatic metrics are stable at smaller sample sizes, only some automatic metrics are able to moderately predict model win rates according to human preference.

Computation and Language

Advancing Generative AI for Portuguese with Open Decoder Gerv\'asio PT*

To advance the neural decoding of Portuguese, in this paper we present a fully open Transformer-based, instruction-tuned decoder model that sets a new state of the art in this respect. To develop this decoder, which we named Gerv\'asio PT*, a strong LLaMA~2 7B model was used as a starting point, and its further improvement through additional training was done over language resources that include new instruction data sets of Portuguese prepared for this purpose, which are also contributed in this paper. All versions of Gerv\'asio are open source and distributed for free under an open license, including for either research or commercial usage, and can be run on consumer-grade hardware, thus seeking to contribute to the advancement of research and innovation in language technology for Portuguese.

Computation and Language

On the Decision-Making Abilities in Role-Playing using Large Language Models

Large language models (LLMs) are now increasingly utilized for role-playing tasks, especially in impersonating domain-specific experts, primarily through role-playing prompts. When interacting in real-world scenarios, the decision-making abilities of a role significantly shape its behavioral patterns. In this paper, we concentrate on evaluating the decision-making abilities of LLMs post role-playing thereby validating the efficacy of role-playing. Our goal is to provide metrics and guidance for enhancing the decision-making abilities of LLMs in role-playing tasks. Specifically, we first use LLMs to generate virtual role descriptions corresponding to the 16 personality types of Myers-Briggs Type Indicator (abbreviated as MBTI) representing a segmentation of the population. Then we design specific quantitative operations to evaluate the decision-making abilities of LLMs post role-playing from four aspects: adaptability, exploration$\&$exploitation trade-off ability, reasoning ability, and safety. Finally, we analyze the association between the performance of decision-making and the corresponding MBTI types through GPT-4. Extensive experiments demonstrate stable differences in the four aspects of decision-making abilities across distinct roles, signifying a robust correlation between decision-making abilities and the roles emulated by LLMs. These results underscore that LLMs can effectively impersonate varied roles while embodying their genuine sociological characteristics.

Computation and Language

How do Large Language Models Handle Multilingualism?

Large language models (LLMs) demonstrate remarkable performance across a spectrum of languages. In this work, we delve into the question: How do LLMs handle multilingualism? We introduce a framework that depicts LLMs' processing of multilingual inputs: In the first several layers, LLMs understand the question, converting multilingual inputs into English to facilitate the task-solving phase. In the intermediate layers, LLMs engage in problem-solving by thinking in English and incorporating multilingual knowledge to obtain factual content, leveraging the self-attention and feed-forward structures, respectively. In the last several layers, LLMs generate responses that align with the original language of the query. In addition, we investigate the existence of language-specific neurons when processing a certain language. To detect neurons activated by the input language, even without labels, we innovatively design a Parallel Language specific Neuron Detection ($\texttt{PLND}$) method that effectively measures the significance of neurons when handling multilingual inputs. By comprehensive ablation analysis through deactivating neurons of different layers and structures, we verify the framework that we propose. Additionally, we demonstrate that we can utilize such a framework to effectively enhance the multilingual ability with much less training effort.

Computation and Language

Utilizing Local Hierarchy with Adversarial Training for Hierarchical Text Classification

Hierarchical text classification (HTC) is a challenging subtask of multi-label classification due to its complex taxonomic structure. Nearly all recent HTC works focus on how the labels are structured but ignore the sub-structure of ground-truth labels according to each input text which contains fruitful label co-occurrence information. In this work, we introduce this local hierarchy with an adversarial framework. We propose a HiAdv framework that can fit in nearly all HTC models and optimize them with the local hierarchy as auxiliary information. We test on two typical HTC models and find that HiAdv is effective in all scenarios and is adept at dealing with complex taxonomic hierarchies. Further experiments demonstrate that the promotion of our framework indeed comes from the local hierarchy and the local hierarchy is beneficial for rare classes which have insufficient training data.

Computation and Language

When does word order matter and when doesn't it?

Language models (LMs) may appear insensitive to word order changes in natural language understanding (NLU) tasks. In this paper, we propose that linguistic redundancy can explain this phenomenon, whereby word order and other linguistic cues such as case markers provide overlapping and thus redundant information. Our hypothesis is that models exhibit insensitivity to word order when the order provides redundant information, and the degree of insensitivity varies across tasks. We quantify how informative word order is using mutual information (MI) between unscrambled and scrambled sentences. Our results show the effect that the less informative word order is, the more consistent the model's predictions are between unscrambled and scrambled sentences. We also find that the effect varies across tasks: for some tasks, like SST-2, LMs' prediction is almost always consistent with the original one even if the Pointwise-MI (PMI) changes, while for others, like RTE, the consistency is near random when the PMI gets lower, i.e., word order is really important.

Computation and Language

Reducing Hallucinations in Entity Abstract Summarization with Facts-Template Decomposition

Entity abstract summarization aims to generate a coherent description of a given entity based on a set of relevant Internet documents. Pretrained language models (PLMs) have achieved significant success in this task, but they may suffer from hallucinations, i.e. generating non-factual information about the entity. To address this issue, we decompose the summary into two components: Facts that represent the factual information about the given entity, which PLMs are prone to fabricate; and Template that comprises generic content with designated slots for facts, which PLMs can generate competently. Based on the facts-template decomposition, we propose SlotSum, an explainable framework for entity abstract summarization. SlotSum first creates the template and then predicts the fact for each template slot based on the input documents. Benefiting from our facts-template decomposition, SlotSum can easily locate errors and further rectify hallucinated predictions with external knowledge. We construct a new dataset WikiFactSum to evaluate the performance of SlotSum. Experimental results demonstrate that SlotSum could generate summaries that are significantly more factual with credible external knowledge.

Computation and Language

Principal Component Analysis as a Sanity Check for Bayesian Phylolinguistic Reconstruction

Bayesian approaches to reconstructing the evolutionary history of languages rely on the tree model, which assumes that these languages descended from a common ancestor and underwent modifications over time. However, this assumption can be violated to different extents due to contact and other factors. Understanding the degree to which this assumption is violated is crucial for validating the accuracy of phylolinguistic inference. In this paper, we propose a simple sanity check: projecting a reconstructed tree onto a space generated by principal component analysis. By using both synthetic and real data, we demonstrate that our method effectively visualizes anomalies, particularly in the form of jogging.

Computation and Language

Updating Language Models with Unstructured Facts: Towards Practical Knowledge Editing

Knowledge editing aims to inject knowledge updates into language models to keep them correct and up-to-date. However, its current evaluation strategies are notably impractical: they solely update with well-curated structured facts (triplets with subjects, relations, and objects), whereas real-world knowledge updates commonly emerge in unstructured texts like news articles. In this paper, we propose a new benchmark, Unstructured Knowledge Editing (UKE). It evaluates editing performance directly using unstructured texts as knowledge updates, termed unstructured facts. Hence UKE avoids the laborious construction of structured facts and enables efficient and responsive knowledge editing, becoming a more practical benchmark. We conduct extensive experiments on newly built datasets and demonstrate that UKE poses a significant challenge to state-of-the-art knowledge editing methods, resulting in their critical performance declines. We further show that this challenge persists even if we extract triplets as structured facts. Our analysis discloses key insights to motivate future research in UKE for more practical knowledge editing.

Computation and Language

AdaMergeX: Cross-Lingual Transfer with Large Language Models via Adaptive Adapter Merging

As an effective alternative to the direct fine-tuning on target tasks in specific languages, cross-lingual transfer addresses the challenges of limited training data by decoupling ''task ability'' and ''language ability'' by fine-tuning on the target task in the source language and another selected task in the target language, respectively. However, they fail to fully separate the task ability from the source language or the language ability from the chosen task. In this paper, we acknowledge the mutual reliance between task ability and language ability and direct our attention toward the gap between the target language and the source language on tasks. As the gap removes the impact of tasks, we assume that it remains consistent across tasks. Based on this assumption, we propose a new cross-lingual transfer method called $\texttt{AdaMergeX}$ that utilizes adaptive adapter merging. By introducing a reference task, we can determine that the divergence of adapters fine-tuned on the reference task in both languages follows the same distribution as the divergence of adapters fine-tuned on the target task in both languages. Hence, we can obtain target adapters by combining the other three adapters. Furthermore, we propose a structure-adaptive adapter merging method. Our empirical results demonstrate that our approach yields new and effective cross-lingual transfer, outperforming existing methods across all settings.

Computation and Language

Inappropriate Pause Detection In Dysarthric Speech Using Large-Scale Speech Recognition

Dysarthria, a common issue among stroke patients, severely impacts speech intelligibility. Inappropriate pauses are crucial indicators in severity assessment and speech-language therapy. We propose to extend a large-scale speech recognition model for inappropriate pause detection in dysarthric speech. To this end, we propose task design, labeling strategy, and a speech recognition model with an inappropriate pause prediction layer. First, we treat pause detection as speech recognition, using an automatic speech recognition (ASR) model to convert speech into text with pause tags. According to the newly designed task, we label pause locations at the text level and their appropriateness. We collaborate with speech-language pathologists to establish labeling criteria, ensuring high-quality annotated data. Finally, we extend the ASR model with an inappropriate pause prediction layer for end-to-end inappropriate pause detection. Moreover, we propose a task-tailored metric for evaluating inappropriate pause detection independent of ASR performance. Our experiments show that the proposed method better detects inappropriate pauses in dysarthric speech than baselines. (Inappropriate Pause Error Rate: 14.47%)

Computation and Language

SemEval 2024 -- Task 10: Emotion Discovery and Reasoning its Flip in Conversation (EDiReF)

We present SemEval-2024 Task 10, a shared task centred on identifying emotions and finding the rationale behind their flips within monolingual English and Hindi-English code-mixed dialogues. This task comprises three distinct subtasks - emotion recognition in conversation for code-mixed dialogues, emotion flip reasoning for code-mixed dialogues, and emotion flip reasoning for English dialogues. Participating systems were tasked to automatically execute one or more of these subtasks. The datasets for these tasks comprise manually annotated conversations focusing on emotions and triggers for emotion shifts (The task data is available at https://github.com/LCS2-IIITD/EDiReF-SemEval2024.git). A total of 84 participants engaged in this task, with the most adept systems attaining F1-scores of 0.70, 0.79, and 0.76 for the respective subtasks. This paper summarises the results and findings from 24 teams alongside their system descriptions.

Computation and Language

PopALM: Popularity-Aligned Language Models for Social Media Trendy Response Prediction

Social media platforms are daily exhibiting millions of events. To preliminarily predict the mainstream public reaction to these events, we study trendy response prediction to automatically generate top-liked user replies to social media events. While previous works focus on generating responses without factoring in popularity, we propose Popularity-Aligned Language Models (PopALM) to distinguish responses liked by a larger audience through reinforcement learning. Recognizing the noisy labels from user "likes", we tailor-make curriculum learning in proximal policy optimization (PPO) to help models capture the essential samples for easy-to-hard training. In experiments, we build a large-scale Weibo dataset for trendy response prediction, and its results show that PopALM can help boost the performance of advanced language models.

Computation and Language

Exploring the Efficacy of Large Language Models in Summarizing Mental Health Counseling Sessions: A Benchmark Study

Comprehensive summaries of sessions enable an effective continuity in mental health counseling, facilitating informed therapy planning. Yet, manual summarization presents a significant challenge, diverting experts' attention from the core counseling process. This study evaluates the effectiveness of state-of-the-art Large Language Models (LLMs) in selectively summarizing various components of therapy sessions through aspect-based summarization, aiming to benchmark their performance. We introduce MentalCLOUDS, a counseling-component guided summarization dataset consisting of 191 counseling sessions with summaries focused on three distinct counseling components (aka counseling aspects). Additionally, we assess the capabilities of 11 state-of-the-art LLMs in addressing the task of component-guided summarization in counseling. The generated summaries are evaluated quantitatively using standard summarization metrics and verified qualitatively by mental health professionals. Our findings demonstrate the superior performance of task-specific LLMs such as MentalLlama, Mistral, and MentalBART in terms of standard quantitative metrics such as Rouge-1, Rouge-2, Rouge-L, and BERTScore across all aspects of counseling components. Further, expert evaluation reveals that Mistral supersedes both MentalLlama and MentalBART based on six parameters -- affective attitude, burden, ethicality, coherence, opportunity costs, and perceived effectiveness. However, these models share the same weakness by demonstrating a potential for improvement in the opportunity costs and perceived effectiveness metrics.

Computation and Language

Pointing out the Shortcomings of Relation Extraction Models with Semantically Motivated Adversarials

In recent years, large language models have achieved state-of-the-art performance across various NLP tasks. However, investigations have shown that these models tend to rely on shortcut features, leading to inaccurate predictions and causing the models to be unreliable at generalization to out-of-distribution (OOD) samples. For instance, in the context of relation extraction (RE), we would expect a model to identify the same relation independently of the entities involved in it. For example, consider the sentence "Leonardo da Vinci painted the Mona Lisa" expressing the created(Leonardo_da_Vinci, Mona_Lisa) relation. If we substiute "Leonardo da Vinci" with "Barack Obama", then the sentence still expresses the created relation. A robust model is supposed to detect the same relation in both cases. In this work, we describe several semantically-motivated strategies to generate adversarial examples by replacing entity mentions and investigate how state-of-the-art RE models perform under pressure. Our analyses show that the performance of these models significantly deteriorates on the modified datasets (avg. of -48.5% in F1), which indicates that these models rely to a great extent on shortcuts, such as surface forms (or patterns therein) of entities, without making full use of the information present in the sentences.

Computation and Language

Controllable Preference Optimization: Toward Controllable Multi-Objective Alignment

Alignment in artificial intelligence pursues the consistency between model responses and human preferences as well as values. In practice, the multifaceted nature of human preferences inadvertently introduces what is known as the "alignment tax" -a compromise where enhancements in alignment within one objective (e.g.,harmlessness) can diminish performance in others (e.g.,helpfulness). However, existing alignment techniques are mostly unidirectional, leading to suboptimal trade-offs and poor flexibility over various objectives. To navigate this challenge, we argue the prominence of grounding LLMs with evident preferences. We introduce controllable preference optimization (CPO), which explicitly specifies preference scores for different objectives, thereby guiding the model to generate responses that meet the requirements. Our experimental analysis reveals that the aligned models can provide responses that match various preferences among the "3H" (helpfulness, honesty, harmlessness) desiderata. Furthermore, by introducing diverse data and alignment goals, we surpass baseline methods in aligning with single objectives, hence mitigating the impact of the alignment tax and achieving Pareto improvements in multi-objective alignment.

Computation and Language

Survey in Characterization of Semantic Change

Live languages continuously evolve to integrate the cultural change of human societies. This evolution manifests through neologisms (new words) or \textbf{semantic changes} of words (new meaning to existing words). Understanding the meaning of words is vital for interpreting texts coming from different cultures (regionalism or slang), domains (e.g., technical terms), or periods. In computer science, these words are relevant to computational linguistics algorithms such as translation, information retrieval, question answering, etc. Semantic changes can potentially impact the quality of the outcomes of these algorithms. Therefore, it is important to understand and characterize these changes formally. The study of this impact is a recent problem that has attracted the attention of the computational linguistics community. Several approaches propose methods to detect semantic changes with good precision, but more effort is needed to characterize how the meaning of words changes and to reason about how to reduce the impact of semantic change. This survey provides an understandable overview of existing approaches to the \textit{characterization of semantic changes} and also formally defines three classes of characterizations: if the meaning of a word becomes more general or narrow (change in dimension) if the word is used in a more pejorative or positive/ameliorated sense (change in orientation), and if there is a trend to use the word in a, for instance, metaphoric or metonymic context (change in relation). We summarized the main aspects of the selected publications in a table and discussed the needs and trends in the research activities on semantic change characterization.

Computation and Language

TEncDM: Understanding the Properties of Diffusion Model in the Space of Language Model Encodings

Drawing inspiration from the success of diffusion models in various domains, numerous research papers proposed methods for adapting them to text data. Despite these efforts, none of them has managed to achieve the quality of the large language models. In this paper, we conduct a comprehensive analysis of key components of the text diffusion models and introduce a novel approach named Text Encoding Diffusion Model (TEncDM). Instead of the commonly used token embedding space, we train our model in the space of the language model encodings. Additionally, we propose to use a Transformer-based decoder that utilizes contextual information for text reconstruction. We also analyse self-conditioning and find that it increases the magnitude of the model outputs, allowing the reduction of the number of denoising steps at the inference stage. Evaluation of TEncDM on two downstream text generation tasks, QQP and XSum, demonstrates its superiority over existing non-autoregressive models.

Computation and Language

Whispers that Shake Foundations: Analyzing and Mitigating False Premise Hallucinations in Large Language Models

Large Language Models (LLMs) have shown impressive capabilities but still suffer from the issue of hallucinations. A significant type of this issue is the false premise hallucination, which we define as the phenomenon when LLMs generate hallucinated text when confronted with false premise questions. In this paper, we perform a comprehensive analysis of the false premise hallucination and elucidate its internal working mechanism: a small subset of attention heads (which we designate as false premise heads) disturb the knowledge extraction process, leading to the occurrence of false premise hallucination. Based on our analysis, we propose \textbf{FAITH} (\textbf{F}alse premise \textbf{A}ttention head constra\textbf{I}ining for mi\textbf{T}igating \textbf{H}allucinations), a novel and effective method to mitigate false premise hallucinations. It constrains the false premise attention heads during the model inference process. Impressively, extensive experiments demonstrate that constraining only approximately $1\%$ of the attention heads in the model yields a notable increase of nearly $20\%$ of model performance.

Computation and Language

How to Understand "Support"? An Implicit-enhanced Causal Inference Approach for Weakly-supervised Phrase Grounding

Weakly-supervised Phrase Grounding (WPG) is an emerging task of inferring the fine-grained phrase-region matching, while merely leveraging the coarse-grained sentence-image pairs for training. However, existing studies on WPG largely ignore the implicit phrase-region matching relations, which are crucial for evaluating the capability of models in understanding the deep multimodal semantics. To this end, this paper proposes an Implicit-Enhanced Causal Inference (IECI) approach to address the challenges of modeling the implicit relations and highlighting them beyond the explicit. Specifically, this approach leverages both the intervention and counterfactual techniques to tackle the above two challenges respectively. Furthermore, a high-quality implicit-enhanced dataset is annotated to evaluate IECI and detailed evaluations show the great advantages of IECI over the state-of-the-art baselines. Particularly, we observe an interesting finding that IECI outperforms the advanced multimodal LLMs by a large margin on this implicit-enhanced dataset, which may facilitate more research to evaluate the multimodal LLMs in this direction.

Computation and Language

Evaluating Webcam-based Gaze Data as an Alternative for Human Rationale Annotations

Rationales in the form of manually annotated input spans usually serve as ground truth when evaluating explainability methods in NLP. They are, however, time-consuming and often biased by the annotation process. In this paper, we debate whether human gaze, in the form of webcam-based eye-tracking recordings, poses a valid alternative when evaluating importance scores. We evaluate the additional information provided by gaze data, such as total reading times, gaze entropy, and decoding accuracy with respect to human rationale annotations. We compare WebQAmGaze, a multilingual dataset for information-seeking QA, with attention and explainability-based importance scores for 4 different multilingual Transformer-based language models (mBERT, distil-mBERT, XLMR, and XLMR-L) and 3 languages (English, Spanish, and German). Our pipeline can easily be applied to other tasks and languages. Our findings suggest that gaze data offers valuable linguistic insights that could be leveraged to infer task difficulty and further show a comparable ranking of explainability methods to that of human rationales.

Computation and Language

Teaching Large Language Models an Unseen Language on the Fly

Existing large language models struggle to support numerous low-resource languages, particularly the extremely low-resource ones where there is minimal training data available for effective parameter updating. We thus investigate whether LLMs can learn a new language on the fly solely through prompting. To study this question, we collect a research suite for Zhuang, a language supported by no LLMs currently. We introduce \textsc{DiPMT++}, a framework for adapting LLMs to unseen languages by in-context learning. Using a dictionary and only 5K parallel sentences, \textsc{DiPMT++} significantly enhances the performance of GPT-4 from 0 to 16 BLEU for Chinese-to-Zhuang translation and achieves 32 BLEU for Zhuang-to-Chinese translation. Furthermore, we demonstrate the practical utility of this framework in aiding humans to translate completely unseen languages, which could contribute to the preservation of linguistic diversity.

Computation and Language

Improving Legal Judgement Prediction in Romanian with Long Text Encoders

In recent years,the entire field of Natural Language Processing (NLP) has enjoyed amazing novel results achieving almost human-like performance on a variety of tasks. Legal NLP domain has also been part of this process, as it has seen an impressive growth. However, general-purpose models are not readily applicable for legal domain. Due to the nature of the domain (e.g. specialized vocabulary, long documents) specific models and methods are often needed for Legal NLP. In this work we investigate both specialized and general models for predicting the final ruling of a legal case, task known as Legal Judgment Prediction (LJP). We particularly focus on methods to extend to sequence length of Transformer-based models to better understand the long documents present in legal corpora. Extensive experiments on 4 LJP datasets in Romanian, originating from 2 sources with significantly different sizes and document lengths, show that specialized models and handling long texts are critical for a good performance.

Computation and Language

PeLLE: Encoder-based language models for Brazilian Portuguese based on open data

In this paper we present PeLLE, a family of large language models based on the RoBERTa architecture, for Brazilian Portuguese, trained on curated, open data from the Carolina corpus. Aiming at reproducible results, we describe details of the pretraining of the models. We also evaluate PeLLE models against a set of existing multilingual and PT-BR refined pretrained Transformer-based LLM encoders, contrasting performance of large versus smaller-but-curated pretrained models in several downstream tasks. We conclude that several tasks perform better with larger models, but some tasks benefit from smaller-but-curated data in its pretraining.

Computation and Language

Memory-Augmented Generative Adversarial Transformers

Conversational AI systems that rely on Large Language Models, like Transformers, have difficulty interweaving external data (like facts) with the language they generate. Vanilla Transformer architectures are not designed for answering factual questions with high accuracy. This paper investigates a possible route for addressing this problem. We propose to extend the standard Transformer architecture with an additional memory bank holding extra information (such as facts drawn from a knowledge base), and an extra attention layer for addressing this memory. We add this augmented memory to a Generative Adversarial Network-inspired Transformer architecture. This setup allows for implementing arbitrary felicity conditions on the generated language of the Transformer. We first demonstrate how this machinery can be deployed for handling factual questions in goal-oriented dialogues. Secondly, we demonstrate that our approach can be useful for applications like {\it style adaptation} as well: the adaptation of utterances according to certain stylistic (external) constraints, like social properties of human interlocutors in dialogues.

Computation and Language

Let LLMs Take on the Latest Challenges! A Chinese Dynamic Question Answering Benchmark

How to better evaluate the capabilities of Large Language Models (LLMs) is the focal point and hot topic in current LLMs research. Previous work has noted that due to the extremely high cost of iterative updates of LLMs, they are often unable to answer the latest dynamic questions well. To promote the improvement of Chinese LLMs' ability to answer dynamic questions, in this paper, we introduce CDQA, a Chinese Dynamic QA benchmark containing question-answer pairs related to the latest news on the Chinese Internet. We obtain high-quality data through a pipeline that combines humans and models, and carefully classify the samples according to the frequency of answer changes to facilitate a more fine-grained observation of LLMs' capabilities. We have also evaluated and analyzed mainstream and advanced Chinese LLMs on CDQA. Extensive experiments and valuable insights suggest that our proposed CDQA is challenging and worthy of more further study. We believe that the benchmark we provide will become one of the key data resources for improving LLMs' Chinese question-answering ability in the future.

Computation and Language

GSM-Plus: A Comprehensive Benchmark for Evaluating the Robustness of LLMs as Mathematical Problem Solvers

Large language models (LLMs) have achieved impressive performance across various mathematical reasoning benchmarks. However, there are increasing debates regarding whether these models truly understand and apply mathematical knowledge or merely rely on shortcuts for mathematical reasoning. One essential and frequently occurring evidence is that when the math questions are slightly changed, LLMs can behave incorrectly. This motivates us to evaluate the robustness of LLMs' math reasoning capability by testing a wide range of question variations. We introduce the adversarial grade school math (\datasetname) dataset, an extension of GSM8K augmented with various mathematical perturbations. Our experiments on 25 LLMs and 4 prompting techniques show that while LLMs exhibit different levels of math reasoning abilities, their performances are far from robust. In particular, even for problems that have been solved in GSM8K, LLMs can make mistakes when new statements are added or the question targets are altered. We also explore whether more robust performance can be achieved by composing existing prompting methods, in which we try an iterative method that generates and verifies each intermediate thought based on its reasoning goal and calculation result. Code and data are available at \url{https://github.com/qtli/GSM-Plus}.

Computation and Language

Robust Guidance for Unsupervised Data Selection: Capturing Perplexing Named Entities for Domain-Specific Machine Translation

Employing extensive datasets enables the training of multilingual machine translation models; however, these models often fail to accurately translate sentences within specialized domains. Although obtaining and translating domain-specific data incurs high costs, it is inevitable for high-quality translations. Hence, finding the most 'effective' data with an unsupervised setting becomes a practical strategy for reducing labeling costs. Recent research indicates that this effective data could be found by selecting 'properly difficult data' based on its volume. This means the data should not be excessively challenging or overly simplistic, especially if the amount of data is limited. However, we found that establishing a criterion for unsupervised data selection remains challenging, as the 'proper difficulty' might vary based on the data domain being trained on. We introduce a novel unsupervised data selection method, 'Capturing Perplexing Named Entities', which adopts the maximum inference entropy in translated named entities as a selection measure. The motivation was that named entities in domain-specific data are considered the most complex portion of the data and should be predicted with high confidence. When verified with the 'Korean-English Parallel Corpus of Specialized Domains,' our method served as a robust guidance for unsupervised data selection, in contrast to existing methods.

Computation and Language

PlanGPT: Enhancing Urban Planning with Tailored Language Model and Efficient Retrieval

In the field of urban planning, general-purpose large language models often struggle to meet the specific needs of planners. Tasks like generating urban planning texts, retrieving related information, and evaluating planning documents pose unique challenges. To enhance the efficiency of urban professionals and overcome these obstacles, we introduce PlanGPT, the first specialized Large Language Model tailored for urban and spatial planning. Developed through collaborative efforts with institutions like the Chinese Academy of Urban Planning, PlanGPT leverages a customized local database retrieval framework, domain-specific fine-tuning of base models, and advanced tooling capabilities. Empirical tests demonstrate that PlanGPT has achieved advanced performance, delivering responses of superior quality precisely tailored to the intricacies of urban planning.

Computation and Language

WanJuan-CC: A Safe and High-Quality Open-sourced English Webtext Dataset

This paper presents WanJuan-CC, a safe and high-quality open-sourced English webtext dataset derived from Common Crawl data. The study addresses the challenges of constructing large-scale pre-training datasets for language models, which require vast amounts of high-quality data. A comprehensive process was designed to handle Common Crawl data, including extraction, heuristic rule filtering, fuzzy deduplication, content safety filtering, and data quality filtering. From approximately 68 billion original English documents, we obtained 2.22T Tokens of safe data and selected 1.0T Tokens of high-quality data as part of WanJuan-CC. We have open-sourced 100B Tokens from this dataset. The paper also provides statistical information related to data quality, enabling users to select appropriate data according to their needs. To evaluate the quality and utility of the dataset, we trained 1B-parameter and 3B-parameter models using WanJuan-CC and another dataset, RefinedWeb. Results show that WanJuan-CC performs better on validation datasets and downstream tasks.

Computation and Language

Compact Speech Translation Models via Discrete Speech Units Pretraining

Using Self-Supervised Learning (SSL) as model initialization is now common to obtain strong results in Speech Translation (ST). However, they also impose a large memory footprint, hindering on-device deployment. In this paper, we leverage the SSL models by pretraining smaller models on their Discrete Speech Units (DSU). We pretrain encoder-decoder models on 1) Filterbank-to-DSU and 2) DSU-to-Translation data, and take the encoder from 1) and the decoder from 2) to initialise a new model, finetuning this on limited speech-translation data. The final model becomes compact by using the DSU pretraining to distil the knowledge of the SSL model. Our method has several benefits over using DSU as model inputs, such as shorter inference pipeline and robustness over (DSU) tokenization. In contrast to ASR pretraining, it does not require transcripts, making it applicable to low-resource settings. Evaluation on CoVoST-2 X-En shows that our method is >$0.5$ BLEU better than a ST model that directly finetune the SSL model, given only half the model size, and on a par with ASR pretraining.

Computation and Language

Here's a Free Lunch: Sanitizing Backdoored Models with Model Merge

The democratization of pre-trained language models through open-source initiatives has rapidly advanced innovation and expanded access to cutting-edge technologies. However, this openness also brings significant security risks, including backdoor attacks, where hidden malicious behaviors are triggered by specific inputs, compromising natural language processing (NLP) system integrity and reliability. This paper suggests that merging a backdoored model with other homogeneous models can remediate backdoor vulnerabilities even if such models are not entirely secure. In our experiments, we explore various models (BERT-Base, RoBERTa-Large, Llama2-7B, and Mistral-7B) and datasets (SST-2, OLID, AG News, and QNLI). Compared to multiple advanced defensive approaches, our method offers an effective and efficient inference-stage defense against backdoor attacks without additional resources or specific knowledge. Our approach consistently outperforms the other advanced baselines, leading to an average of 75% reduction in the attack success rate. Since model merging has been an established approach for improving model performance, the extra advantage it provides regarding defense can be seen as a cost-free bonus.

Computation and Language

Prompting Explicit and Implicit Knowledge for Multi-hop Question Answering Based on Human Reading Process

Pre-trained language models (PLMs) leverage chains-of-thought (CoT) to simulate human reasoning and inference processes, achieving proficient performance in multi-hop QA. However, a gap persists between PLMs' reasoning abilities and those of humans when tackling complex problems. Psychological studies suggest a vital connection between explicit information in passages and human prior knowledge during reading. Nevertheless, current research has given insufficient attention to linking input passages and PLMs' pre-training-based knowledge from the perspective of human cognition studies. In this study, we introduce a Prompting Explicit and Implicit knowledge (PEI) framework, which uses prompts to connect explicit and implicit knowledge, aligning with human reading process for multi-hop QA. We consider the input passages as explicit knowledge, employing them to elicit implicit knowledge through unified prompt reasoning. Furthermore, our model incorporates type-specific reasoning via prompts, a form of implicit knowledge. Experimental results show that PEI performs comparably to the state-of-the-art on HotpotQA. Ablation studies confirm the efficacy of our model in bridging and integrating explicit and implicit knowledge.

Computation and Language

OpenMedLM: Prompt engineering can out-perform fine-tuning in medical question-answering with open-source large language models

LLMs have become increasingly capable at accomplishing a range of specialized-tasks and can be utilized to expand equitable access to medical knowledge. Most medical LLMs have involved extensive fine-tuning, leveraging specialized medical data and significant, thus costly, amounts of computational power. Many of the top performing LLMs are proprietary and their access is limited to very few research groups. However, open-source (OS) models represent a key area of growth for medical LLMs due to significant improvements in performance and an inherent ability to provide the transparency and compliance required in healthcare. We present OpenMedLM, a prompting platform which delivers state-of-the-art (SOTA) performance for OS LLMs on medical benchmarks. We evaluated a range of OS foundation LLMs (7B-70B) on four medical benchmarks (MedQA, MedMCQA, PubMedQA, MMLU medical-subset). We employed a series of prompting strategies, including zero-shot, few-shot, chain-of-thought (random selection and kNN selection), and ensemble/self-consistency voting. We found that OpenMedLM delivers OS SOTA results on three common medical LLM benchmarks, surpassing the previous best performing OS models that leveraged computationally costly extensive fine-tuning. The model delivers a 72.6% accuracy on the MedQA benchmark, outperforming the previous SOTA by 2.4%, and achieves 81.7% accuracy on the MMLU medical-subset, establishing itself as the first OS LLM to surpass 80% accuracy on this benchmark. Our results highlight medical-specific emergent properties in OS LLMs which have not yet been documented to date elsewhere, and showcase the benefits of further leveraging prompt engineering to improve the performance of accessible LLMs for medical applications.

Computation and Language

On the Scaling Laws of Geographical Representation in Language Models

Language models have long been shown to embed geographical information in their hidden representations. This line of work has recently been revisited by extending this result to Large Language Models (LLMs). In this paper, we propose to fill the gap between well-established and recent literature by observing how geographical knowledge evolves when scaling language models. We show that geographical knowledge is observable even for tiny models, and that it scales consistently as we increase the model size. Notably, we observe that larger language models cannot mitigate the geographical bias that is inherent to the training data.

Computation and Language

$\texttt{COSMIC}$: Mutual Information for Task-Agnostic Summarization Evaluation

Assessing the quality of summarizers poses significant challenges. In response, we propose a novel task-oriented evaluation approach that assesses summarizers based on their capacity to produce summaries that are useful for downstream tasks, while preserving task outcomes. We theoretically establish a direct relationship between the resulting error probability of these tasks and the mutual information between source texts and generated summaries. We introduce $\texttt{COSMIC}$ as a practical implementation of this metric, demonstrating its strong correlation with human judgment-based metrics and its effectiveness in predicting downstream task performance. Comparative analyses against established metrics like $\texttt{BERTScore}$ and $\texttt{ROUGE}$ highlight the competitive performance of $\texttt{COSMIC}$.

Computation and Language

Towards Tracing Trustworthiness Dynamics: Revisiting Pre-training Period of Large Language Models

Ensuring the trustworthiness of large language models (LLMs) is crucial. Most studies concentrate on fully pre-trained LLMs to better understand and improve LLMs' trustworthiness. In this paper, to reveal the untapped potential of pre-training, we pioneer the exploration of LLMs' trustworthiness during this period, focusing on five key dimensions: reliability, privacy, toxicity, fairness, and robustness. To begin with, we apply linear probing to LLMs. The high probing accuracy suggests that \textit{LLMs in early pre-training can already distinguish concepts in each trustworthiness dimension}. Therefore, to further uncover the hidden possibilities of pre-training, we extract steering vectors from a LLM's pre-training checkpoints to enhance the LLM's trustworthiness. Finally, inspired by~\citet{choi2023understanding} that mutual information estimation is bounded by linear probing accuracy, we also probe LLMs with mutual information to investigate the dynamics of trustworthiness during pre-training. We are the first to observe a similar two-phase phenomenon: fitting and compression~\citep{shwartz2017opening}. This research provides an initial exploration of trustworthiness modeling during LLM pre-training, seeking to unveil new insights and spur further developments in the field. We will make our code publicly accessible at \url{https://github.com/ChnQ/TracingLLM}.

Computation and Language

TV-TREES: Multimodal Entailment Trees for Neuro-Symbolic Video Reasoning

It is challenging to perform question-answering over complex, multimodal content such as television clips. This is in part because current video-language models rely on single-modality reasoning, have lowered performance on long inputs, and lack interpetability. We propose TV-TREES, the first multimodal entailment tree generator. TV-TREES serves as an approach to video understanding that promotes interpretable joint-modality reasoning by producing trees of entailment relationships between simple premises directly entailed by the videos and higher-level conclusions. We then introduce the task of multimodal entailment tree generation to evaluate the reasoning quality of such methods. Our method's experimental results on the challenging TVQA dataset demonstrate intepretable, state-of-the-art zero-shot performance on full video clips, illustrating a best of both worlds contrast to black-box methods.

Computation and Language

Loose LIPS Sink Ships: Asking Questions in Battleship with Language-Informed Program Sampling

Questions combine our mastery of language with our remarkable facility for reasoning about uncertainty. How do people navigate vast hypothesis spaces to pose informative questions given limited cognitive resources? We study these tradeoffs in a classic grounded question-asking task based on the board game Battleship. Our language-informed program sampling (LIPS) model uses large language models (LLMs) to generate natural language questions, translate them into symbolic programs, and evaluate their expected information gain. We find that with a surprisingly modest resource budget, this simple Monte Carlo optimization strategy yields informative questions that mirror human performance across varied Battleship board scenarios. In contrast, LLM-only baselines struggle to ground questions in the board state; notably, GPT-4V provides no improvement over non-visual baselines. Our results illustrate how Bayesian models of question-asking can leverage the statistics of language to capture human priors, while highlighting some shortcomings of pure LLMs as grounded reasoners.

Computation and Language

Query-OPT: Optimizing Inference of Large Language Models via Multi-Query Instructions in Meeting Summarization

This work focuses on the task of query-based meeting summarization in which the summary of a context (meeting transcript) is generated in response to a specific query. When using Large Language Models (LLMs) for this task, a new call to the LLM inference endpoint/API is required for each new query even if the context stays the same. However, repeated calls to the LLM inference endpoints would significantly increase the costs of using them in production, making LLMs impractical for many real-world use cases. To address this problem, in this paper, we investigate whether combining the queries for the same input context in a single prompt to minimize repeated calls can be successfully used in meeting summarization. In this regard, we conduct extensive experiments by comparing the performance of various popular LLMs: GPT-4, PaLM-2, LLaMA-2, Mistral, and FLAN-T5 in single-query and multi-query settings. We observe that while most LLMs tend to respond to the multi-query instructions, almost all of them (except GPT-4), even after fine-tuning, could not properly generate the response in the required output format. We conclude that while multi-query prompting could be useful to optimize the inference costs by reducing calls to the inference endpoints/APIs for the task of meeting summarization, this capability to reliably generate the response in the expected format is only limited to certain LLMs.

Computation and Language

Resonance RoPE: Improving Context Length Generalization of Large Language Models

This paper addresses the challenge of train-short-test-long (TSTL) scenarios in Large Language Models (LLMs) equipped with Rotary Position Embedding (RoPE), where models pre-trained on shorter sequences face difficulty with out-of-distribution (OOD) token positions in longer sequences. We introduce Resonance RoPE, a novel approach designed to narrow the generalization gap in TSTL scenarios by refining the interpolation of RoPE features for OOD positions, significantly improving the model performance without additional online computational costs. Furthermore, we present PosGen, a new synthetic benchmark specifically designed for fine-grained behavior analysis in TSTL scenarios, aiming to isolate the constantly increasing difficulty of token generation on long contexts from the challenges of recognizing new token positions. Our experiments on synthetic tasks show that after applying Resonance RoPE, Transformers recognize OOD position better and more robustly. Our extensive LLM experiments also show superior performance after applying Resonance RoPE to the current state-of-the-art RoPE scaling method, YaRN, on both upstream language modeling tasks and a variety of downstream long-text applications.

Computation and Language

PROC2PDDL: Open-Domain Planning Representations from Texts

Planning in a text-based environment continues to be a major challenge for AI systems. Recent approaches have used language models to predict a planning domain definition (e.g., PDDL) but have only been evaluated in closed-domain simulated environments. To address this, we present Proc2PDDL , the first dataset containing open-domain procedural texts paired with expert-annotated PDDL representations. Using this dataset, we evaluate state-of-the-art models on defining the preconditions and effects of actions. We show that Proc2PDDL is highly challenging, with GPT-3.5's success rate close to 0% and GPT-4's around 35%. Our analysis shows both syntactic and semantic errors, indicating LMs' deficiency in both generating domain-specific prgorams and reasoning about events. We hope this analysis and dataset helps future progress towards integrating the best of LMs and formal planning.

Computation and Language

FAC$^2$E: Better Understanding Large Language Model Capabilities by Dissociating Language and Cognition

Large language models (LLMs) are primarily evaluated by overall performance on various text understanding and generation tasks. However, such a paradigm fails to comprehensively differentiate the fine-grained language and cognitive skills, rendering the lack of sufficient interpretation to LLMs' capabilities. In this paper, we present FAC$^2$E, a framework for Fine-grAined and Cognition-grounded LLMs' Capability Evaluation. Specifically, we formulate LLMs' evaluation in a multi-dimensional and explainable manner by dissociating the language-related capabilities and the cognition-related ones. Besides, through extracting the intermediate reasoning from LLMs, we further break down the process of applying a specific capability into three sub-steps: recalling relevant knowledge, utilizing knowledge, and solving problems. Finally, FAC$^2$E evaluates each sub-step of each fine-grained capability, providing a two-faceted diagnosis for LLMs. Utilizing FAC$^2$E, we identify a common shortfall in knowledge utilization among models and propose a straightforward, knowledge-enhanced method to mitigate this issue. Our results not only showcase promising performance enhancements but also highlight a direction for future LLM advancements.

Computation and Language

Prompting ChatGPT for Translation: A Comparative Analysis of Translation Brief and Persona Prompts

Prompt engineering in LLMs has shown potential for improving translation quality. However, the potential of incorporating translation concepts in prompt design remains largely underexplored. Against this backdrop, this paper discusses the effectiveness of incorporating the conceptual tool of translation brief and the personas of translator and author into prompt design for translation tasks in ChatGPT. Findings suggest that, although certain elements are constructive in facilitating human to human communication for translation tasks, their effectiveness is limited for improving translation quality in ChatGPT. This accentuates the need for more explorative research on how translation theorists and practitioners can develop the current set of conceptual tools rooted in the human to human communication paradigm for translation purposes in this emerging workflow involving human machine interaction.

Computation and Language

EROS: Entity-Driven Controlled Policy Document Summarization

Privacy policy documents have a crucial role in educating individuals about the collection, usage, and protection of users' personal data by organizations. However, they are notorious for their lengthy, complex, and convoluted language especially involving privacy-related entities. Hence, they pose a significant challenge to users who attempt to comprehend organization's data usage policy. In this paper, we propose to enhance the interpretability and readability of policy documents by using controlled abstractive summarization -- we enforce the generated summaries to include critical privacy-related entities (e.g., data and medium) and organization's rationale (e.g.,target and reason) in collecting those entities. To achieve this, we develop PD-Sum, a policy-document summarization dataset with marked privacy-related entity labels. Our proposed model, EROS, identifies critical entities through a span-based entity extraction model and employs them to control the information content of the summaries using proximal policy optimization (PPO). Comparison shows encouraging improvement over various baselines. Furthermore, we furnish qualitative and human evaluations to establish the efficacy of EROS.

Computation and Language

Ensemble-Based Unsupervised Discontinuous Constituency Parsing by Tree Averaging

We address unsupervised discontinuous constituency parsing, where we observe a high variance in the performance of the only previous model. We propose to build an ensemble of different runs of the existing discontinuous parser by averaging the predicted trees, to stabilize and boost performance. To begin with, we provide comprehensive computational complexity analysis (in terms of P and NP-complete) for tree averaging under different setups of binarity and continuity. We then develop an efficient exact algorithm to tackle the task, which runs in a reasonable time for all samples in our experiments. Results on three datasets show our method outperforms all baselines in all metrics; we also provide in-depth analyses of our approach.

Computation and Language

EBBS: An Ensemble with Bi-Level Beam Search for Zero-Shot Machine Translation

The ability of zero-shot translation emerges when we train a multilingual model with certain translation directions; the model can then directly translate in unseen directions. Alternatively, zero-shot translation can be accomplished by pivoting through a third language (e.g., English). In our work, we observe that both direct and pivot translations are noisy and achieve less satisfactory performance. We propose EBBS, an ensemble method with a novel bi-level beam search algorithm, where each ensemble component explores its own prediction step by step at the lower level but they are synchronized by a "soft voting" mechanism at the upper level. Results on two popular multilingual translation datasets show that EBBS consistently outperforms direct and pivot translations as well as existing ensemble techniques. Further, we can distill the ensemble's knowledge back to the multilingual model to improve inference efficiency; profoundly, our EBBS-based distillation does not sacrifice, or even improves, the translation quality.

Computation and Language

TELEClass: Taxonomy Enrichment and LLM-Enhanced Hierarchical Text Classification with Minimal Supervision

Hierarchical text classification aims to categorize each document into a set of classes in a label taxonomy. Most earlier works focus on fully or semi-supervised methods that require a large amount of human annotated data which is costly and time-consuming to acquire. To alleviate human efforts, in this paper, we work on hierarchical text classification with the minimal amount of supervision: using the sole class name of each node as the only supervision. Recently, large language models (LLM) show competitive performance on various tasks through zero-shot prompting, but this method performs poorly in the hierarchical setting, because it is ineffective to include the large and structured label space in a prompt. On the other hand, previous weakly-supervised hierarchical text classification methods only utilize the raw taxonomy skeleton and ignore the rich information hidden in the text corpus that can serve as additional class-indicative features. To tackle the above challenges, we propose TELEClass, Taxonomy Enrichment and LLM-Enhanced weakly-supervised hierarchical text classification, which (1) automatically enriches the label taxonomy with class-indicative topical terms mined from the corpus to facilitate classifier training and (2) utilizes LLMs for both data annotation and creation tailored for the hierarchical label space. Experiments show that TELEClass can outperform previous weakly-supervised hierarchical text classification methods and LLM-based zero-shot prompting methods on two public datasets.

Computation and Language

"Flex Tape Can't Fix That": Bias and Misinformation in Edited Language Models

Model editing has emerged as a cost-effective strategy to update knowledge stored in language models. However, model editing can have unintended consequences after edits are applied: information unrelated to the edits can also be changed, and other general behaviors of the model can be wrongly altered. In this work, we investigate how model editing methods unexpectedly amplify model biases post-edit. We introduce a novel benchmark dataset, Seesaw-CF, for measuring bias-related harms of model editing and conduct the first in-depth investigation of how different weight-editing methods impact model bias. Specifically, we focus on biases with respect to demographic attributes such as race, geographic origin, and gender, as well as qualitative flaws in long-form texts generated by edited language models. We find that edited models exhibit, to various degrees, more biased behavior as they become less confident in attributes for Asian, African, and South American subjects. Furthermore, edited models amplify sexism and xenophobia in text generations while remaining seemingly coherent and logical. Finally, editing facts about place of birth, country of citizenship, or gender have particularly negative effects on the model's knowledge about unrelated features like field of work.

Computation and Language

AXOLOTL: Fairness through Assisted Self-Debiasing of Large Language Model Outputs

Pre-trained Large Language Models (LLMs) have significantly advanced natural language processing capabilities but are susceptible to biases present in their training data, leading to unfair outcomes in various applications. While numerous strategies have been proposed to mitigate bias, they often require extensive computational resources and may compromise model performance. In this work, we introduce AXOLOTL, a novel post-processing framework, which operates agnostically across tasks and models, leveraging public APIs to interact with LLMs without direct access to internal parameters. Through a three-step process resembling zero-shot learning, AXOLOTL identifies biases, proposes resolutions, and guides the model to self-debias its outputs. This approach minimizes computational costs and preserves model performance, making AXOLOTL a promising tool for debiasing LLM outputs with broad applicability and ease of use.

Computation and Language

Improving Socratic Question Generation using Data Augmentation and Preference Optimization

The Socratic method is a way of guiding students toward solving a problem independently without directly revealing the solution to the problem. Although this method has been shown to significantly improve student learning outcomes, it remains a complex labor-intensive task for instructors. Large language models (LLMs) can be used to augment human effort by automatically generating Socratic questions for students. However, existing methods that involve prompting these LLMs sometimes produce invalid outputs, e.g., those that directly reveal the solution to the problem or provide irrelevant or premature questions. To alleviate this problem, inspired by reinforcement learning with AI feedback (RLAIF), we first propose a data augmentation method to enrich existing Socratic questioning datasets with questions that are invalid in specific ways. Next, we propose a method to optimize open-source LLMs such as LLama 2 to prefer ground-truth questions over generated invalid ones, using direct preference optimization (DPO). Our experiments on a Socratic questions dataset for student code debugging show that a DPO-optimized 7B LLama 2 model can effectively avoid generating invalid questions, and as a result, outperforms existing state-of-the-art prompting methods.

Computation and Language

Transcription and translation of videos using fine-tuned XLSR Wav2Vec2 on custom dataset and mBART

This research addresses the challenge of training an ASR model for personalized voices with minimal data. Utilizing just 14 minutes of custom audio from a YouTube video, we employ Retrieval-Based Voice Conversion (RVC) to create a custom Common Voice 16.0 corpus. Subsequently, a Cross-lingual Self-supervised Representations (XLSR) Wav2Vec2 model is fine-tuned on this dataset. The developed web-based GUI efficiently transcribes and translates input Hindi videos. By integrating XLSR Wav2Vec2 and mBART, the system aligns the translated text with the video timeline, delivering an accessible solution for multilingual video content transcription and translation for personalized voice.

Computation and Language