Papers arxiv:2605.13941 EvolveMem:Self-Evolving Memory Architecture via AutoResearch for LLM Agents Published on May 13 Submitted by JiaqiLiu on May 15 Authors: Jiaqi Liu , , , , , , Abstract EvolveMem enables adaptive memory…
…At the first level, the Harness Evolution Loop optimizes a worker agent 's harness H for a single task: a Worker Agent W_{H} executes the task, an Evaluator Agent V adversarially…
…open-source agent by 9.9% in accuracy with 5.3x fewer tool-call rounds on average. View arXiv page View PDF GitHub 49 Add to collection Community Code Repo: https://github…
…Skill-Triggered Audit for Request-Conditioned Invocation Safety in Agent Systems (2026) CodeTracer: Towards Traceable Agent States (2026) Schema First Tool APIs for LLM Agents: A Controlled Study of Tool Misuse, Recovery…
…benchmarks spanning Mathematics, Code, Agent Tasks, and Function Calling. Notably, ResRL surpasses NSR on mathematical reasoning by 9.4\% in Avg@16 and 7.0\% in Pass@128. Code is available at…
…Adaptive Entropy Modulation for Multi-Turn Agentic Reinforcement Learning (2026) DPEPO: Diverse Parallel Exploration Policy Optimization for LLM-based Agents (2026) GRPO-VPS: Enhancing Group Relative Policy Optimization with Verifiable Process Supervision…
…Agentic Autoregressive Diffusion for Long Video Consistency Published on May 7 Submitted by Do Xuan Long on May 11 Google Authors: , , , , Abstract A$^2$RD, an Agentic Auto-Regressive Diffusion architecture, addresses…
…Advancing PRMs for Reinforcing Code Agents (2026) ClawEnvKit: Automatic Environment Generation for Claw-Like Agents (2026) ClawGUI: A Unified Framework for Training, Evaluating, and Deploying GUI Agents (2026) SWE-Next: Scalable Real…
…Our code is available at https://github.com/Trustworthy-ML-Lab/when2tool View arXiv page View PDF Project page GitHub 5 Add to collection Community LLM agents often call tools even when…
…y}) I can see the Docker image amhma/ubuntu-desktop:22.04-0.0.1-dev is used, but the server code (if that lives there) and Dockerfile would that would be…