hassenhamdi
hassenhamdi
AI & ML interests
None yet
Recent Activity
commented on
a paper
1 day ago
Scalable Power Sampling: Unlocking Efficient, Training-Free Reasoning for LLMs via Distribution Sharpening
commented on
a paper
1 day ago
Scalable Power Sampling: Unlocking Efficient, Training-Free Reasoning for LLMs via Distribution Sharpening
updated
a collection
1 day ago
Optimization
Organizations
Coding
Diffusion
MLLM
-
NextFlow: Unified Sequential Modeling Activates Multimodal Understanding and Generation
Paper ā¢ 2601.02204 ā¢ Published ā¢ 61 -
HERMES: KV Cache as Hierarchical Memory for Efficient Streaming Video Understanding
Paper ā¢ 2601.14724 ā¢ Published ā¢ 73 -
VIOLA: Towards Video In-Context Learning with Minimal Annotations
Paper ā¢ 2601.15549 ā¢ Published ā¢ 4
Agents
-
Imagine-then-Plan: Agent Learning from Adaptive Lookahead with World Models
Paper ā¢ 2601.08955 ā¢ Published ā¢ 13 -
EvoFSM: Controllable Self-Evolution for Deep Research with Finite State Machines
Paper ā¢ 2601.09465 ā¢ Published ā¢ 40 -
MAXS: Meta-Adaptive Exploration with LLM Agents
Paper ā¢ 2601.09259 ā¢ Published ā¢ 95 -
Toward Efficient Agents: Memory, Tool learning, and Planning
Paper ā¢ 2601.14192 ā¢ Published ā¢ 51
LLM
-
A Safety Report on GPT-5.2, Gemini 3 Pro, Qwen3-VL, Doubao 1.8, Grok 4.1 Fast, Nano Banana Pro, and Seedream 4.5
Paper ā¢ 2601.10527 ā¢ Published ā¢ 24 -
PACEvolve: Enabling Long-Horizon Progress-Aware Consistent Evolution
Paper ā¢ 2601.10657 ā¢ Published ā¢ 20 -
TranslateGemma Technical Report
Paper ā¢ 2601.09012 ā¢ Published ā¢ 19 -
Recursive Language Models
Paper ā¢ 2512.24601 ā¢ Published ā¢ 78
RAG
-
From RAG to Memory: Non-Parametric Continual Learning for Large Language Models
Paper ā¢ 2502.14802 ā¢ Published ā¢ 13 -
EvoFSM: Controllable Self-Evolution for Deep Research with Finite State Machines
Paper ā¢ 2601.09465 ā¢ Published ā¢ 40 -
Parallel Context-of-Experts Decoding for Retrieval Augmented Generation
Paper ā¢ 2601.08670 ā¢ Published ā¢ 19
Architecture
Training-Free
-
HERMES: KV Cache as Hierarchical Memory for Efficient Streaming Video Understanding
Paper ā¢ 2601.14724 ā¢ Published ā¢ 73 -
LLM-in-Sandbox Elicits General Agentic Intelligence
Paper ā¢ 2601.16206 ā¢ Published ā¢ 82 -
Scalable Power Sampling: Unlocking Efficient, Training-Free Reasoning for LLMs via Distribution Sharpening
Paper ā¢ 2601.21590 ā¢ Published ā¢ 11
Optimization
-
MHLA: Restoring Expressivity of Linear Attention via Token-Level Multi-Head
Paper ā¢ 2601.07832 ā¢ Published ā¢ 51 -
Towards Automated Kernel Generation in the Era of LLMs
Paper ā¢ 2601.15727 ā¢ Published ā¢ 16 -
HERMES: KV Cache as Hierarchical Memory for Efficient Streaming Video Understanding
Paper ā¢ 2601.14724 ā¢ Published ā¢ 73 -
Post-LayerNorm Is Back: Stable, ExpressivE, and Deep
Paper ā¢ 2601.19895 ā¢ Published ā¢ 19
Video generation
Algorithmic Evolution
-
PACEvolve: Enabling Long-Horizon Progress-Aware Consistent Evolution
Paper ā¢ 2601.10657 ā¢ Published ā¢ 20 -
Controlled Self-Evolution for Algorithmic Code Optimization
Paper ā¢ 2601.07348 ā¢ Published ā¢ 113 -
LoongFlow: Directed Evolutionary Search via a Cognitive Plan-Execute-Summarize Paradigm
Paper ā¢ 2512.24077 ā¢ Published ā¢ 4
Papers
Vision
Architecture
Coding
Training-Free
-
HERMES: KV Cache as Hierarchical Memory for Efficient Streaming Video Understanding
Paper ā¢ 2601.14724 ā¢ Published ā¢ 73 -
LLM-in-Sandbox Elicits General Agentic Intelligence
Paper ā¢ 2601.16206 ā¢ Published ā¢ 82 -
Scalable Power Sampling: Unlocking Efficient, Training-Free Reasoning for LLMs via Distribution Sharpening
Paper ā¢ 2601.21590 ā¢ Published ā¢ 11
Diffusion
Optimization
-
MHLA: Restoring Expressivity of Linear Attention via Token-Level Multi-Head
Paper ā¢ 2601.07832 ā¢ Published ā¢ 51 -
Towards Automated Kernel Generation in the Era of LLMs
Paper ā¢ 2601.15727 ā¢ Published ā¢ 16 -
HERMES: KV Cache as Hierarchical Memory for Efficient Streaming Video Understanding
Paper ā¢ 2601.14724 ā¢ Published ā¢ 73 -
Post-LayerNorm Is Back: Stable, ExpressivE, and Deep
Paper ā¢ 2601.19895 ā¢ Published ā¢ 19
MLLM
-
NextFlow: Unified Sequential Modeling Activates Multimodal Understanding and Generation
Paper ā¢ 2601.02204 ā¢ Published ā¢ 61 -
HERMES: KV Cache as Hierarchical Memory for Efficient Streaming Video Understanding
Paper ā¢ 2601.14724 ā¢ Published ā¢ 73 -
VIOLA: Towards Video In-Context Learning with Minimal Annotations
Paper ā¢ 2601.15549 ā¢ Published ā¢ 4
Video generation
Agents
-
Imagine-then-Plan: Agent Learning from Adaptive Lookahead with World Models
Paper ā¢ 2601.08955 ā¢ Published ā¢ 13 -
EvoFSM: Controllable Self-Evolution for Deep Research with Finite State Machines
Paper ā¢ 2601.09465 ā¢ Published ā¢ 40 -
MAXS: Meta-Adaptive Exploration with LLM Agents
Paper ā¢ 2601.09259 ā¢ Published ā¢ 95 -
Toward Efficient Agents: Memory, Tool learning, and Planning
Paper ā¢ 2601.14192 ā¢ Published ā¢ 51
Algorithmic Evolution
-
PACEvolve: Enabling Long-Horizon Progress-Aware Consistent Evolution
Paper ā¢ 2601.10657 ā¢ Published ā¢ 20 -
Controlled Self-Evolution for Algorithmic Code Optimization
Paper ā¢ 2601.07348 ā¢ Published ā¢ 113 -
LoongFlow: Directed Evolutionary Search via a Cognitive Plan-Execute-Summarize Paradigm
Paper ā¢ 2512.24077 ā¢ Published ā¢ 4
LLM
-
A Safety Report on GPT-5.2, Gemini 3 Pro, Qwen3-VL, Doubao 1.8, Grok 4.1 Fast, Nano Banana Pro, and Seedream 4.5
Paper ā¢ 2601.10527 ā¢ Published ā¢ 24 -
PACEvolve: Enabling Long-Horizon Progress-Aware Consistent Evolution
Paper ā¢ 2601.10657 ā¢ Published ā¢ 20 -
TranslateGemma Technical Report
Paper ā¢ 2601.09012 ā¢ Published ā¢ 19 -
Recursive Language Models
Paper ā¢ 2512.24601 ā¢ Published ā¢ 78
Papers
RAG
-
From RAG to Memory: Non-Parametric Continual Learning for Large Language Models
Paper ā¢ 2502.14802 ā¢ Published ā¢ 13 -
EvoFSM: Controllable Self-Evolution for Deep Research with Finite State Machines
Paper ā¢ 2601.09465 ā¢ Published ā¢ 40 -
Parallel Context-of-Experts Decoding for Retrieval Augmented Generation
Paper ā¢ 2601.08670 ā¢ Published ā¢ 19