SlopCodeBench: Benchmarking How Coding Agents Degrade over Long, Iterative Tasks

2 pointsposted 13 hours ago
by softwaredoug
(arxiv.org)

No comments yet