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

arXiv: https://arxiv.org/abs/2506.19724 · HTML: https://mlsanity.com/paper/2506.19724/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- Placed in Web, search, and research agents, one of 16 directions in LLM agents and tool use.
- Builds on 23 of its 41 references that are in the map.
- First posted on arXiv 24 Jun 2025.

## Placed in

- LLM agents and tool use / [Web, search, and research agents](https://mlsanity.com/agents/web-search-and-research-agents/)

## Builds on (most cited first, 23 of 23)

- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023)
- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [Do As I Can, Not As I Say: Grounding Language in Robotic Affordances](https://mlsanity.com/paper/2204.01691/) (2022)
- [SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering](https://mlsanity.com/paper/2405.15793/) (2024)
- [The Rise and Potential of Large Language Model Based Agents: A Survey](https://mlsanity.com/paper/2309.07864/) (2023)
- [s1: Simple test-time scaling](https://mlsanity.com/paper/2501.19393/) (2025)
- [The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery](https://mlsanity.com/paper/2408.06292/) (2024)
- [Measuring Coding Challenge Competence With APPS](https://mlsanity.com/paper/2105.09938/) (2021)
- [LLM-Planner: Few-Shot Grounded Planning for Embodied Agents with Large Language Models](https://mlsanity.com/paper/2212.04088/) (2022)
- [MLAgentBench: Evaluating Language Agents on Machine Learning Experimentation](https://mlsanity.com/paper/2310.03302/) (2023)
- [Selection-Inference: Exploiting Large Language Models for Interpretable Logical Reasoning](https://mlsanity.com/paper/2205.09712/) (2022)
- [PaperBench: Evaluating AI's Ability to Replicate AI Research](https://mlsanity.com/paper/2504.01848/) (2025)
- [Chain of Agents: Large Language Models Collaborating on Long-Context Tasks](https://mlsanity.com/paper/2406.02818/) (2024)
- [SciMON: Scientific Inspiration Machines Optimized for Novelty](https://mlsanity.com/paper/2305.14259/) (2023)
- [ReviewerGPT? An Exploratory Study on Using Large Language Models for Paper Reviewing](https://mlsanity.com/paper/2306.00622/) (2023)
- [Natural Language to Code Generation in Interactive Data Science Notebooks](https://mlsanity.com/paper/2212.09248/) (2022)
- [Beyond the Limits: A Survey of Techniques to Extend the Context Length in Large Language Models](https://mlsanity.com/paper/2402.02244/) (2024)
- [TroVE: Inducing Verifiable and Efficient Toolboxes for Solving Programmatic Tasks](https://mlsanity.com/paper/2401.12869/) (2024)
- [Large Language Models are Zero Shot Hypothesis Proposers](https://mlsanity.com/paper/2311.05965/) (2023)
- [AI-Driven Review Systems: Evaluating LLMs in Scalable and Bias-Aware Academic Reviews](https://mlsanity.com/paper/2408.10365/) (2024)
- [Sustainable computational science: the ReScience initiative](https://mlsanity.com/paper/1707.04393/) (2017)

## Built on by (most cited first, 12 of 12)

- [NatureBench: Can Coding Agents Match the Published SOTA of Nature-Family Papers?](https://mlsanity.com/paper/2606.24530/) (2026)
- [Training AI Scientists to Replicate Research](https://mlsanity.com/paper/2608.13331/) (2026)
- [HiRAS: A Hierarchical Multi-Agent Framework for Paper-to-Code Generation and Execution](https://mlsanity.com/paper/2604.17745/) (2026)
- [The Implementation Lottery: Auditing Idea Reliability in Automated Research](https://mlsanity.com/paper/2607.26587/) (2026)
- [From paper to benchmark: agentic, framework-based reproduction of under-specified methods in machine health intelligence](https://mlsanity.com/paper/2605.28371/) (2026)
- [Towards Autonomous and Auditable Medical Imaging Model Development](https://mlsanity.com/paper/2607.10522/) (2026)
- [Overview of the NLPCC 2026 Shared Task 11: Agent-Based Experiment Reproduction from Scientific Papers](https://mlsanity.com/paper/2609.11117/) (2026)
- [ARA: Agentic Reproducibility Assessment For Scalable Support Of Scientific Peer-Review](https://mlsanity.com/paper/2605.02651/) (2026)
- [Terminal Agents: A Survey of AI Agents in Command-Line Environments](https://mlsanity.com/paper/2608.20485/) (2026)
- [ArtifactLinker: Linking Scientific Artifacts for Automatic State-of-the-Art Discovery](https://mlsanity.com/paper/2605.16902/) (2026)
- [RECLAIM: Can Agents Reproduce the Claims of Machine Learning Papers?](https://mlsanity.com/paper/2609.28850/) (2026)
- [Agents Are Systems, Not Models: Rethinking Agentic Evaluation](https://mlsanity.com/paper/2610.01618/) (2026)

## Abstract (arXiv)

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 .