Pith. sign in

REVIEW 10 cited by

From Reproduction to Replication: Evaluating Research Agents with Progressive Code Masking

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2506.19724 v1 pith:BIUOSSU6 submitted 2025-06-24 cs.AI

classification cs.AI
keywords codeagentsautoexperimentbenchmarkexperimentreplicationreproductionresearch
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
abstract

Recent progress in autonomous code generation has fueled excitement around AI agents capable of accelerating scientific discovery by running experiments. However, there is currently no benchmark that evaluates whether such agents can implement scientific ideas when given varied amounts of code as a starting point, interpolating between reproduction (running code) and from-scratch replication (fully re-implementing and running code). We introduce AutoExperiment, a benchmark that evaluates AI agents' ability to implement and run machine learning experiments based on natural language descriptions in research papers. In each task, agents are given a research paper, a codebase with key functions masked out, and a command to run the experiment. The goal is to generate the missing code, execute the experiment in a sandboxed environment, and reproduce the results. AutoExperiment scales in difficulty by varying the number of missing functions $n$, ranging from partial reproduction to full replication. We evaluate state-of-the-art agents and find that performance degrades rapidly as $n$ increases. Agents that can dynamically interact with the environment (e.g. to debug their code) can outperform agents in fixed "agentless" harnesses, and there exists a significant gap between single-shot and multi-trial success rates (Pass@1 vs. Pass@5), motivating verifier approaches to our benchmark. Our findings highlight critical challenges in long-horizon code generation, context retrieval, and autonomous experiment execution, establishing AutoExperiment as a new benchmark for evaluating progress in AI-driven scientific experimentation. Our data and code are open-sourced at https://github.com/j1mk1m/AutoExperiment .

Discussion (0). Sign in to comment.

Forward citations

Cited by 10 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. One Run Is Not an Idea: The Implementation Lottery in Automated Research

    cs.MA 2026-07 conditional novelty 7.0 of 10

    In audits of 312 LLM research assignments, variation across implementations of the same idea dominates rerun noise and reverses the selected idea winner in 25.6-43.6% of decisions.

  2. Coding-agents can replicate scientific machine learning papers

    cs.AI 2026-07 unverdicted novelty 7.0 of 10

    Paper-replication is a workflow that enables coding agents to replicate computational claims from scientific ML papers by recording targets, reconstructing methods, running experiments, and validating evidence against...

  3. NatureBench: Can Coding Agents Match the Published SOTA of Nature-Family Papers?

    cs.CL 2026-06 accept novelty 7.0 of 10

    Frontier coding agents surpass published Nature-family SOTA on only 17.8% of 90 sealed scientific tasks, mostly by recasting problems as supervised ML rather than inventing methods.

  4. NatureBench: Can Coding Agents Match the Published SOTA of Nature-Family Papers?

    cs.CL 2026-06 unverdicted novelty 7.0 of 10

    NatureBench evaluates ten frontier AI coding agents on 90 tasks from Nature papers under web-search-disabled conditions and finds the strongest agent surpasses published SOTA on only 17.8% of tasks, succeeding mainly ...

  5. Towards Autonomous and Auditable Medical Imaging Model Development

    cs.CV 2026-07 conditional novelty 6.0 of 10

    AMID, a verification-guided multi-agent MLE system for medical imaging, outperforms general MLE agents on 20 ReX-MLE challenges and approaches human challenge solutions on several tasks.

  6. Coding-agents can replicate scientific machine learning papers

    cs.AI 2026-07 conditional novelty 6.0 of 10

    Paper-replication makes coding agents complete SciML paper replications only when every recorded claim has provenance, comparison evidence, and report coverage in a validated workspace.

  7. From paper to benchmark: agentic, framework-based reproduction of under-specified methods in machine health intelligence

    cs.AI 2026-05 unverdicted novelty 6.0 of 10

    Proposes agentic framework-based reproduction with a slot-binding interface to turn 16 PHM papers into standardized, assumption-aware benchmark implementations.

  8. ArtifactLinker: Linking Scientific Artifacts for Automatic State-of-the-Art Discovery

    cs.LG 2026-05 unverdicted novelty 6.0 of 10

    ArtifactLinker frames SOTA discovery as missing-link prediction on an artifact graph of models and datasets, with a two-stage ranking-plus-verification pipeline and a new benchmark of 14k artifacts.

  9. ARA: Agentic Reproducibility Assessment For Scalable Support Of Scientific Peer-Review

    cs.DL 2026-05 unverdicted novelty 6.0 of 10

    ARA uses LLMs to build workflow graphs linking sources, methods, and outputs in papers, then scores reproducibility, reaching ~61% accuracy on 213 ReScience C articles and outperforming priors on ReproBench and GoldSt...

  10. ARA: Agentic Reproducibility Assessment For Scalable Support Of Scientific Peer-Review

    cs.DL 2026-05 unverdicted novelty 6.0 of 10

    ARA extracts workflow graphs from papers and scores reproducibility, reaching 61% accuracy on 213 ReScience C articles and outperforming priors on ReproBench and GoldStandardDB.

Pith tools