| ▲ | Show HN: OmnisBench, a re-gradable, open LLM routing benchmark on fresh tasks(github.com) | |
| 3 points by fortitudedev 5 hours ago | 1 comments | ||
| ▲ | quietraster 4 hours ago | parent [-] | |
fresh tasks plus regrading could make routing benchmarks much harder to game. how often will the task set rotate? | ||