Paper 2605.20025
AutoResearchClaw: Self-Reinforcing Autonomous Research with Human-AI Collaboration
- Published
- May 2026
- Research lab
- Independent
- Citations
- 13
- GitHub
- 14K stars
01 In brief
Summary
AutoResearchClaw is a multi-agent autonomous research pipeline that addresses three key challenges in scientific discovery: hypothesis quality, execution robustness, and experience accumulation.
It integrates five mechanisms: structured multi-agent debate for hypothesis generation and result analysis, a self-healing executor with a Pivot/Refine loop, verifiable result reporting with a numeric registry and four-layer citation verification, human-in-the-loop collaboration with seven intervention modes, and cross-run evolution using a time-decayed lesson store.
On the new ARC-Bench benchmark (25 ML topics), AutoResearchClaw outperforms AI Scientist v2 by 54.7% and AIDE-ML by 26.8% in experiment-stage evaluation.
A human-in-the-loop ablation across seven modes shows that targeted intervention at high-leverage points (CoPilot) achieves the highest paper quality (7.27 mean score, 87.5% accept rate), outperforming both full autonomy (4.03, 25%) and step-by-step oversight (5.19, 50%).
Component ablation reveals that debate drives quality, self-healing drives completion, and verification ensures integrity, with super-additive effects when combined.
The system also demonstrates cross-domain coverage in physics, biology, and statistics, and is positioned as a research amplifier that augments rather than replaces human judgment.
02 From the paper
Abstract
Automating scientific discovery requires more than generating papers from ideas. Real research is iterative: hypotheses are challenged from multiple perspectives, experiments fail and inform the next attempt, and lessons accumulate across cycles. Existing autonomous research systems often model this process as a linear pipeline: they rely on single-agent reasoning, stop when execution fails, and do not carry experience across runs. We present AutoResearchClaw, a multi-agent autonomous research pipeline built on five mechanisms: structured multi-agent debate for hypothesis generation and result analysis, a self-healing executor with a \textsc{Pivot}/\textsc{Refine} decision loop that transforms failures into information, verifiable result reporting that prevents fabricated numbers and hallucinated citations, human-in-the-loop collaboration with seven intervention modes spanning full autonomy to step-by-step oversight, and cross-run evolution that converts past mistakes into future safeguards. On ARC-Bench, a 25-topic experiment-stage benchmark, AutoResearchClaw outperforms AI Scientist v2 by 54.7%. A human-in-the-loop ablation across seven intervention modes reveals that precise, targeted collaboration at high-leverage decision points consistently outperforms both full autonomy and exhaustive step-by-step oversight. We position AutoResearchClaw as a research amplifier that augments rather than replaces human scientific judgment. Code is available at https://github.com/aiming-lab/AutoResearchClaw.