SourcesVESPO — Stable Off-Policy LLM Training 152 HF UpvotesarXiv·high signalVariational sequence-level soft policy optimization. Stable training at 64x staleness ratio. Top paper on HuggingFace Feb 23.SourceSource pagearXiv↳ Follow the threadPolicy dependency / Stack layerAria: A General Code Agent + Verification Harness Proves All 4,257 Core Iris Lemmas Fully AutomaticallyarXivPolicy dependency / Stack layerABot-N1: A General Visual-Language Navigation Foundation Model (76 Upvotes on HF Daily)arXiv / HuggingFace Daily PapersPolicy dependency / Stack layerDistributed Backdoors Prove Per-Step Agent Monitors Can Be Right on Every Step and Still Miss the AttackarXivPolicy dependency / Stack layerMulti-agent system does fine-tuning-free clinical symptom detection from unstructured notesarXivPolicy dependency / Stack layerLightMem-Ego: A Lightweight Personal-Memory Layer for Everyday AI Agents Trends on HF DailyarXiv / HuggingFace Daily PapersPolicy dependency / Stack layerMako: A 'Self-Evolving Agentic OS' That Writes New Exploits Against Live Targets and Hot-Loads Them Into ItselfarXivPolicy dependency / Stack layerRetrieval-Oriented Code Representations Sharpen Agentic Bug LocalizationarXivPolicy dependency / Stack layerMCPZoo Study: Scanners Call 96.89% of MCP Servers 'Risky' — But Under Half of Those Alerts Are RealarXiv