DeepSWE (Agentica)
An open RL-trained coding agent, not a benchmark. Its release reports results on SWE-bench Verified with test-based verifiers; listed here for its reproducible evaluation setup.
Failure modes DeepSWE (Agentica) catches
Overfitting to verifiers
Patch errors
Starter command
A starting point, not a guarantee. Commands change between releases, so confirm against the official README.
git clone https://github.com/agentica-project/R2E-Gym
# follow the repo README for setup and run commandsSources
- 1. DeepSWE (Agentica) official page or repository · Agentica; Together AI · Repository · Sources checked