Carlos E. Jimenez

Researcher at Anthropic; co-creator of SWE-Bench and SWE-Agent while a PhD student at Princeton.

Cite

Notes

Only stored in your browser.

Attribution

1papers·2eval contribs

Authored papers

SWE-bench: Can Language Models Resolve Real-World GitHub Issues?

ICLR

OpenAI

500 human-validated SWE-bench tasks confirmed solvable from the issue alone, with non-flaky test suites - the most-reported agentic coding benchmark.

Princeton NLP Group

2,294 real GitHub issues from 12 popular Python repos that require an agent to produce a patch passing the project's test suite.

Currently at

researcher · frontier lab

Previously

from 1 papers

Alexander Wettig

researcher

John Yang

grad-student

Karthik Narasimhan

professor

Kexin Pei

professor

Ofir Press

postdoc

Shunyu Yao

researcher