学术前沿

1253 篇文章

7.0
Shorthand for Thought: Compressing LLM Reasoning via Entropy-Guided Supertokens
2026年05月01日
· 05/01 14:31 采集
6.0
NanoKnow: How to Know What Your Language Model Knows
2026年05月01日
· 05/01 14:31 采集
6.0
Grounding Agent Memory in Contextual Intent
2026年05月01日
· 05/01 14:31 采集
6.0
TiMem: Temporal-Hierarchical Memory Consolidation for Long-Horizon Conversational Agents
2026年05月01日
· 05/01 14:31 采集
6.0
Agentic Memory: Learning Unified Long-Term and Short-Term Memory Management for Large Language Model Agents
2026年05月01日
· 05/01 14:31 采集
7.0
Efficient-DLM: From Autoregressive to Diffusion Language Models, and Beyond in Speed
2026年05月01日
· 05/01 14:31 采集
7.0
Exploration Hacking: Can LLMs Learn to Resist RL Training?
2026年05月01日
· 05/01 14:31 采集
7.0
Latent-GRPO: Group Relative Policy Optimization for Latent Reasoning
2026年05月01日
· 05/01 14:31 采集
6.0
ZipCCL: Efficient Lossless Data Compression of Communication Collectives for Accelerating LLM Training
2026年05月01日
· 05/01 14:31 采集
7.0
How Generative AI Disrupts Search: An Empirical Study of Google Search, Gemini, and AI Overviews
2026年05月01日
· 05/01 14:31 采集
7.0
WindowsWorld: A Process-Centric Benchmark of Autonomous GUI Agents in Professional Cross-Application Environments
2026年05月01日
· 05/01 14:31 采集
7.0
Contextual Agentic Memory is a Memo, Not True Memory
2026年05月01日
· 05/01 14:31 采集
6.0
ScaleBox: Enabling High-Fidelity and Scalable Code Verification for Large Language Models
2026年05月01日
· 05/01 14:31 采集
6.0
To Diff or Not to Diff? Structure-Aware and Adaptive Output Formats for Efficient LLM-based Code Editing
2026年05月01日
· 05/01 14:31 采集
6.0
NanoKnow: How to Know What Your Language Model Knows
2026年05月01日
· 05/01 14:31 采集
7.0
Efficient-DLM: From Autoregressive to Diffusion Language Models, and Beyond in Speed
2026年05月01日
· 05/01 14:31 采集
6.0
Debiasing Reward Models via Causally Motivated Inference-Time Intervention
2026年05月01日
· 05/01 14:31 采集
6.0
Skills-Coach: A Self-Evolving Skill Optimizer via Training-Free GRPO
2026年05月01日
· 05/01 14:31 采集
6.0
Learning When to Remember: Risk-Sensitive Contextual Bandits for Abstention-Aware Memory Retrieval in LLM-Based Coding Agents
2026年05月01日
· 05/01 14:31 采集
7.0
Decoupling the Benefits of Subword Tokenization for Language Model Training via Byte-level Simulation
2026年05月01日
· 05/01 14:31 采集

学术前沿

Shorthand for Thought: Compressing LLM Reasoning via Entropy-Guided Supertokens

NanoKnow: How to Know What Your Language Model Knows

Grounding Agent Memory in Contextual Intent

TiMem: Temporal-Hierarchical Memory Consolidation for Long-Horizon Conversational Agents

Agentic Memory: Learning Unified Long-Term and Short-Term Memory Management for Large Language Model Agents

Efficient-DLM: From Autoregressive to Diffusion Language Models, and Beyond in Speed

Exploration Hacking: Can LLMs Learn to Resist RL Training?

Latent-GRPO: Group Relative Policy Optimization for Latent Reasoning

ZipCCL: Efficient Lossless Data Compression of Communication Collectives for Accelerating LLM Training

How Generative AI Disrupts Search: An Empirical Study of Google Search, Gemini, and AI Overviews

WindowsWorld: A Process-Centric Benchmark of Autonomous GUI Agents in Professional Cross-Application Environments

Contextual Agentic Memory is a Memo, Not True Memory

ScaleBox: Enabling High-Fidelity and Scalable Code Verification for Large Language Models

To Diff or Not to Diff? Structure-Aware and Adaptive Output Formats for Efficient LLM-based Code Editing

NanoKnow: How to Know What Your Language Model Knows

Efficient-DLM: From Autoregressive to Diffusion Language Models, and Beyond in Speed

Debiasing Reward Models via Causally Motivated Inference-Time Intervention

Skills-Coach: A Self-Evolving Skill Optimizer via Training-Free GRPO

Learning When to Remember: Risk-Sensitive Contextual Bandits for Abstention-Aware Memory Retrieval in LLM-Based Coding Agents

Decoupling the Benefits of Subword Tokenization for Language Model Training via Byte-level Simulation