9. Kimi-Dev Kimi-Dev introduces agentless training as a skill prior to software engineering LLMs, bridging workflow-style and agentic paradigms.
@dair_ai
-
Holistic Agent Leaderboard: Standardized AI Agent Evaluation Framework
By
–
10. Holistic Agent Leaderboard The Holistic Agent Leaderboard (HAL) introduces a standardized framework for large-scale, reproducible AI agent evaluation across 9 models and 9 benchmarks, spanning coding, web navigation, science, and customer service.
-
LLMs Brain Rot: Cognitive Degradation from Trivial Web Text
By
–
7. LLMs Can Get “Brain Rot”! The authors test a clear hypothesis: continual pretraining on trivial, highly engaging web text degrades LLM cognition in ways that persist even after mitigation.
-
Hybrid Ensemble Reward Optimization for LLM Reasoning
By
–
8. Hybrid Reinforcement Hybrid Ensemble Reward Optimization is a reinforcement learning framework that combines binary verifier feedback with continuous reward-model signals to improve LLM reasoning.
-
Dynamic Layer Routing: Retrofittable Router Optimization for Frozen LLMs
By
–
6. Dynamic Layer Routing in LLMs A retrofittable way to add per-layer routers to frozen LLMs that decide to skip, execute, or repeat each block.
-
Cell2Sentence-Scale 27B: Converting Gene Expression into AI Training Data
By
–
1. Cell2Sentence-Scale 27B C2S-Scale extends Cell2Sentence by converting gene expression into “cell sentences” and training LLMs on 50M+ cells plus biological text.
-
Top AI Papers of the Week: October 13-19
By
–
Top AI Papers of The Week (October 13-19): – Kimi-Dev
– Elastic-Cache
– Hybrid Reinforcement
– Cell2Sentence-Scale 27B
– Holistic Agent Leaderboard
– Dynamic Layer Routing in LLMs
– The Art of Scaling RL Compute for LLMs Read on for more: -
Webscale-RL: 1.2M QA Pairs for Web-Scale Reinforcement Learning
By
–
10. Webscale-RL Webscale-RL introduces a scalable data pipeline that transforms web-scale pretraining text into over 1.2M diverse, verifiable QA pairs for reinforcement learning across 9+ domains.
-
Artificial Hippocampus Networks: RNN Memory for Transformers
By
–
9. Artificial Hippocampus Networks Adds a fixed-size recurrent memory to sliding-window Transformers, compressing evicted KV into RNN-like states (Mamba2/DN/GDN) trained via self-distillation for long-context efficiency with constant cache and near-linear compute.
-
Mem-Agent: 4B LLM with Persistent Memory via GSPO
By
–
8. mem-agent mem-agent is a 4B-parameter LLM trained with GSPO reinforcement learning to develop persistent memory using a scaffold of Python tools and markdown files.