paper-with-me

Papers

ResearchEnvBench: Benchmarking Agents on Environment Synthesis for Research Code Execution

2026-03-06 · Yubang Wang, Chenxi Zhang, Bowen Chen, Zezheng Huai, Zihao Dai, Xinchi Chen, Yuxin Wang, Yining Zheng, Jingjing Gong, Xipeng Qiu arxiv

Autonomous agents are increasingly expected to support scientific research, and recent benchmarks report progress in code repair and autonomous experimentation. However, these evaluations typically assume a pre-configured execution environment, which requires resolving complex software dependencies, aligning hardware and framework versions, and configuring distributed execution, yet this capability remains largely unbenchmarked. We introduce ResearchEnvBench, a benchmark for environment synthesis in research code execution. Given a research repository, documentation, and a target execution setting, agents must construct an environment that successfully executes at runtime. Evaluations on diverse research repositories reveal a substantial gap in current SOTA agents, with failures dominated by incomplete dependency resolution and brittle version coupling. ResearchEnvBench provides a realistic testbed for advancing autonomous agents toward reproducible scientific research.

📄 PDF Abstract BibTeX arXiv:2603.06739

Code (0)

등록된 구현이 없습니다.

Tasks

Code Repair

Similar Papers 제목 키워드 기반

E-Bench: Benchmarking Multi-Step Tool-Use Agents in Real-World Product Scenarios

2026-07-26 · Weihuang Zheng, Tianyuan Zou, Eileen Ye, Alphet Liu 외 arxiv

Large Language Models (LLMs) are increasingly deployed as agents that interact with stateful environments over multiple steps: gathering hidden information, composing tool calls, and committing state changes. We refer to…

Towards a Playground to Democratize Experimentation and Benchmarking of AI Agents for Network Troubleshooting

2025-07-01 · Zhihao Wang, Alessandro Cornacchia, Franco Galante, Carlo Centofanti 외 arxiv

Recent research has demonstrated the effectiveness of Artificial Intelligence (AI), and more specifically, Large Language Models (LLMs), in supporting network configuration synthesis and automating network diagnosis task…

Gaia2: Benchmarking LLM Agents on Dynamic and Asynchronous Environments

2026-02-12 · Romain Froger, Pierre Andrews, Matteo Bettini, Amar Budhiraja 외 arxiv

We introduce Gaia2, a benchmark for evaluating large language model agents in realistic, asynchronous environments. Unlike prior static or synchronous evaluations, Gaia2 introduces scenarios where environments evolve ind…

Reinforcement Learning

NovelGym: A Flexible Ecosystem for Hybrid Planning and Learning Agents Designed for Open Worlds

2024-01-07 · Shivam Goel, Yichen Wei, Panagiotis Lymperopoulos, Matthias Scheutz 외

As AI agents leave the lab and venture into the real world as autonomous vehicles, delivery robots, and cooking robots, it is increasingly necessary to design and comprehensively evaluate algorithms that tackle the ``ope…

Autonomous VehiclesBenchmarkingReinforcement Learning (RL)

Benchmarking the Spectrum of Agent Capabilities

2021-09-14 · ICLR 2022 4 · Danijar Hafner

Evaluating the general abilities of intelligent agents requires complex simulation environments. Existing benchmarks typically evaluate only one narrow task per environment, requiring researchers to perform expensive tra…

Benchmarking