Hugging Face Daily PapersShuhan Xue, Jianyuan Zhong, Ziyuan NanSep 15, 20261 min readpaperadvanced

ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents

Summary

ScienceBuddy is an interactive workspace that converts researcher prompts, feedback, and execution traces into continual‑learning tasks for AI agents. It introduces a "recursive‑in‑recursive" self‑improvement loop that alternates harness refinement and model training, and showcases case studies across four scientific task families.

ScienceBuddy turns researcher interactions into structured training data for ongoing model improvement.
The core "recursive‑in‑recursive" loop separates harness evolution (fixed model) from model reinforcement learning (updated harness).
Four scientific task families are used as case studies to illustrate the paradigm.
The system and code are released publicly to enable community experimentation.

Researchers building AI assistants for scientific work should care because the paper proposes a systematic way to keep such agents improving through real‑world collaboration.

4/10

llm reinforcement learning artificial intelligence self improvement ai agents

Related reading

ScienceIDE: Turning World's Scientific Codebase into Agent Learnable Environments arxiv.org
ScienceIDE is a framework that turns existing scientific software repositories into programmable environments that agents can use for task generation, execution, and verification. Training on these environments yields LLMs (PhAI‑IDE series) that outperform baselines on scientific code repair and several general code‑reasoning benchmarks.
Hugging Face Daily Papersarxiv.orgHejia Geng, Zesen Huang, Haoyang Li6d2 minpaper
ModularRSI: Modular and Generalizable Recursive Harness Self-Improvement arxiv.org
ModularRSI proposes a contrastive, modular approach to evolve agent harnesses by independently updating five functional modules and aggregating evidence across 2,000 benchmark‑disjoint tasks. Experiments on TB2.0 and SWE‑Bench Verified show consistent gains on unseen tasks and cross‑model transfer.
Hugging Face Daily Papersarxiv.orgSiwei Wu, Jincheng Ren, Yizhi Li8d1 minpaper
Generalized Agent Iteration: One Formal Framework for Iterative Policy Improvement and Recursive Self-Improvement arxiv.org
The paper introduces Generalized Agent Iteration (GAI), a formal framework that unifies classical iterative policy improvement (GPI) and recursive self‑improvement (RSI). GAI treats an agent as a set of modifiable components and models learning as a loop of evaluation and improvement. Two binary “dials”—whether the improvement mechanism is internal to the agent and whether the evaluation standard…
Hugging Face Daily Papersarxiv.orgHongyao Tang, Yi Ma, Pengyi Li11d1 minpaper
The Last AI Built by Humans: Toward Genuine Recursive Self-Improvement arxiv.org
The paper defines recursive self‑improvement (RSI) for AI, introduces the Headroom‑Closed Index to expose current LLM limits, and proposes a staged roadmap toward full RSI. It surveys application scenarios and outlines practical challenges.
Hugging Face Daily Papersarxiv.orgYi Duan, Ying Liu, Zirui Tang12d1 minpaperHN3
RecreationWorld: Scalable and Verifiable Environments for Hybrid Computer-Use Agents arxiv.org
The paper presents RecreationWorld, a five‑platform framework that lets hybrid computer‑use agents learn by recreating the behavior of a running reference, and introduces RecreationBench, a 250‑task benchmark with programmatic and visual assertions. Experiments show GPT‑6 Astra reaches 58.1% overall but struggles with deeper programmatic tests, highlighting gaps in current agents.
Hugging Face Daily Papersarxiv.orgShuai Bai, Jiayong Deng, Yikun Fu4d2 minpaper
SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness arxiv.org
SoL-Pi is a new approach for scaling auto-research loops in coding agents, focusing on token efficiency for long reasoning trajectories. It achieves comparable performance to existing methods while significantly reducing token traffic by 44.7-49.0% and API costs by about one third on the 51-task EdgeBench evaluation.
Hugging Face Daily Papersarxiv.orgHaozhe Liu, Tian Ye, Sensen Gao5d1 minpaper

ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents

Summary

Related reading

ScienceIDE: Turning World's Scientific Codebase into Agent Learnable Environmentsarxiv.org

ModularRSI: Modular and Generalizable Recursive Harness Self-Improvementarxiv.org

Generalized Agent Iteration: One Formal Framework for Iterative Policy Improvement and Recursive Self-Improvementarxiv.org

The Last AI Built by Humans: Toward Genuine Recursive Self-Improvementarxiv.org

RecreationWorld: Scalable and Verifiable Environments for Hybrid Computer-Use Agentsarxiv.org

SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harnessarxiv.org

ScienceIDE: Turning World's Scientific Codebase into Agent Learnable Environments arxiv.org

ModularRSI: Modular and Generalizable Recursive Harness Self-Improvement arxiv.org

Generalized Agent Iteration: One Formal Framework for Iterative Policy Improvement and Recursive Self-Improvement arxiv.org

The Last AI Built by Humans: Toward Genuine Recursive Self-Improvement arxiv.org

RecreationWorld: Scalable and Verifiable Environments for Hybrid Computer-Use Agents arxiv.org

SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness arxiv.org