Reproduce “EnvOpt: Automated Optimization for Agent Environments” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Mingxuan Li, Huanzhi Mao, Qiuyang Mang, Hanchen Li, Joseph E. Gonzalez, Alex Dimakis
EnvOpt optimizes a stationary agent environment by having a coding agent inspect trajectories and edit information, tools, or their organization while keeping the target model and harness fixed. On AutomationBench, Claude Haiku 4.5’s pass rate rose from 2.33% to 16.67%, while dollar cost fell from $42.36 to $11.55 across three runs; this result is limited to the tested setup.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.