DLite: History-free Speculative Decoding for Agentic AI Paper • 2610.23787 • Published 16 days ago • 36
Lexical Reference Frames: Interpreting LM Residual Streams in Native Semantics Paper • 2610.22169 • Published 17 days ago • 56
Terminal-RSI: Recursive Self-Improving Terminal Agents via Recursive Self-Rewrite and Policy-Refine Paper • 2610.22192 • Published 17 days ago • 41
KV4-RL: Unlocking Faster Asynchronous Rollouts for RLVR with INT4 KV Cache Quantization Paper • 2610.14448 • Published 22 days ago • 49
Glance Forcing: Making Bidirectional Diffusion Autoregressive at a Glance Paper • 2610.20220 • Published 18 days ago • 11
Prefill-Free Cross-Family KV Cache Transfer for Heterogeneous Multi-Agent LLMs Paper • 2609.32259 • Published Sep 29 • 104
Video Generation Models: A Survey of Post-Training and Alignment Paper • 2610.00812 • Published Sep 30 • 63
EvoOntology: A Self-Evolving Ontology Layer for Data Agents Paper • 2609.15779 • Published Sep 14 • 131
DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression Paper • 2609.19969 • Published Sep 17 • 233
RetireOPD: Self-Retiring On-Policy Distillation for Agentic Reinforcement Learning Paper • 2609.20784 • Published Sep 17 • 57
NCP-ArchPreview Technical Report: Moving towards Latent Space Language Models through Next Concept Prediction Paper • 2609.10715 • Published Sep 9 • 198
ModularRSI: Modular and Generalizable Recursive Harness Self-Improvement Paper • 2609.14857 • Published Sep 14 • 169
Grouped Value Attention: Efficient KV Caching via On-Demand Key Reconstruction Paper • 2609.13285 • Published Sep 8 • 83
How Lossless Is Lossless Speculative Decoding? The Role of Numerical Precision in Orthrus Paper • 2609.15504 • Published Sep 14 • 37
Occamy-1.0: Open Pareto-frontier 35B Intelligence for Co-work Paper • 2609.11977 • Published Sep 4 • 124
BeaconKV: Key-Value Cache Compression Guided by Beacon Queries for Efficient Large Reasoning Model Inference Paper • 2609.04971 • Published Sep 4 • 36