Zenith
An evaluation harness designed to help normal models compete on software engineering benchmarks.
Why it made the shelf
Zenith is described as a state-of-the-art harness for AI models to tackle software engineering tasks, making it a.
Seen on
- First seen
- 3 months ago
- Best score
- 8 points · 1 comments
Public trail
Where it showed up
1 reference from 1 publisher.
- First trace
- Hacker News3 months ago
- Discussion
- 1 comments
- 01Hacker NewsFirst trace
sota harness for normal models to beat Fable on FrontierSWE
