Predicting Emergent Capabilities by Finetuning
Paper
•
2411.16035
•
Published
•
6
Erwin Weiß
Rexschwert
AI & ML interests
AI, Big Data, Data Science, Machine Learning, Computer Vision, Natural Language Processing
Recent Activity
upvoted
a
paper
14 days ago
Predicting Emergent Capabilities by Finetuning
upvoted
a
paper
14 days ago
LLMs Do Not Think Step-by-step In Implicit Reasoning
upvoted
a
paper
14 days ago
Best of Both Worlds: Advantages of Hybrid Graph Sequence Models
Organizations
Rexschwert's activity
upvoted
7
papers
14 days ago
LLMs Do Not Think Step-by-step In Implicit Reasoning
Paper
•
2411.15862
•
Published
•
8
Best of Both Worlds: Advantages of Hybrid Graph Sequence Models
Paper
•
2411.15671
•
Published
•
7
All Languages Matter: Evaluating LMMs on Culturally Diverse 100 Languages
Paper
•
2411.16508
•
Published
•
7
Knowledge Transfer Across Modalities with Natural Language Supervision
Paper
•
2411.15611
•
Published
•
15
O1 Replication Journey -- Part 2: Surpassing O1-preview through Simple Distillation, Big Progress or Bitter Lesson?
Paper
•
2411.16489
•
Published
•
36
From Generation to Judgment: Opportunities and Challenges of LLM-as-a-judge
Paper
•
2411.16594
•
Published
•
36
upvoted
7
papers
about 1 month ago
Unleashing Reasoning Capability of LLMs via Scalable Question Synthesis from Scratch
Paper
•
2410.18693
•
Published
•
40
Can Knowledge Editing Really Correct Hallucinations?
Paper
•
2410.16251
•
Published
•
54
Breaking the Memory Barrier: Near Infinite Batch Size Scaling for Contrastive Loss
Paper
•
2410.17243
•
Published
•
88
On Memorization of Large Language Models in Logical Reasoning
Paper
•
2410.23123
•
Published
•
17
Stealing User Prompts from Mixture of Experts
Paper
•
2410.22884
•
Published
•
13
A Large Recurrent Action Model: xLSTM enables Fast Inference for Robotics Tasks
Paper
•
2410.22391
•
Published
•
21
Task Vectors are Cross-Modal
Paper
•
2410.22330
•
Published
•
11
upvoted
2
collections
2 months ago