Applied Compute improves Qwen agent efficiency on SWE bench

Applied Compute boosted a Qwen thinking model's submit tool call rate from 22% to 60% on SWE bench, with test pass rate flat, by conditioning rollout on an old production trace. The model now wraps up by turn 40 instead of up to 80.
Featured · Samuel Denton
1 source
Daily brief
Get tomorrow's AI brief in your inbox
More stories today
- GOP panics over Big Tech ties as Trump shifts on AI regulation
- Ethan Mollick: Claude's skill creator beats ChatGPT for reusable skills
- Aident Loadout gives agents 27,000+ tools and logs every action
- Etched gains sizable fan base for AI inferencing computers
- Corbell generates technical specs from repository knowledge graphs