… Here we introduce PaSaMaster, a self-evolving agentic literature retrieval system that produces relevance-scored paper rankings with evidence-grounded recommendations through iterative intent analysis , retrieval, and ranking. It is built on three key designs. …
… SimCoder self-evolves by using verifier feedback e.g., compilation errors, physics checks, VLM critiques to revise environments and autonomously add reusable tools and skills to its library. …
… View arXiv page View PDF Add to collection Community This paper studies self-evolving LLM agents that improve by updating external harnesses, and separates harness-evolution from base task-solving capability along two dimensions: harness-updating, which writes useful persistent updates, and harness… …
Papers arxiv:2605.02943 Healthcare AI GYM for Medical Agents Published on May 1 Submitted by Minbyul Jeong on May 6 Authors: Abstract Multi-turn clinical reasoning training using reinforcement learning reveals degradation into verbose single-turn interactions, which is addressed through a novel sel… …
… First, we construct structure-aware perturbations of expert trajectories along longitudinal progress and lateral heading directions and supervise the model to recover the original expert trajectory. …
… View arXiv page View PDF Project page GitHub 10 Add to collection Community LychSim is a highly controllable, interactive simulation framework built on Unreal Engine 5, designed to lower the technical barrier of using a modern game engine for computer vision research. …
… The following papers were recommended by the Semantic Scholar API Knowledge-Graph Paths as Intermediate Supervision for Self-Evolving Search Agents 2026 CoSearch: Joint Training of Reasoning and Document Ranking via Reinforcement Learning for Agentic Search 2026 IG-Search: Step-Level Information Ga… …
… View arXiv page View PDF Add to collection Community A few threads I'd love to explore with the community: Routing collapse as a design principle — the router rejected RBF entirely and the purpose-built DCT+Attn variant beat all 3-operator versions. …
… View arXiv page View PDF Project page GitHub 2 Add to collection Community LLM agents often need memory to solve tasks in new tool environments, but memory is usually built only after seeing human tasks or deployment-time user interactions. …
… If you want recommendations for any Paper on Hugging Face checkout this Space You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: @ librarian-bot recommend the core move here, turning method-level evolution into a typed graph and steering lineage with Self-Guide… …