01.AI·released 2024-03-011 source
Yi-9B benchmark scores: 4 benchmarks tracked. 0% of its scores are independently reproduced — the rest are self-reported or unverified.
Leads on 0 of 4 benchmarks·best result 68.53 on HellaSwag (unverified)·0 of 4 independently reproduced
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| HellaSwag | 68.53accuracy (%) | unverified· optimizedT1 | 2024-03-01Yi: Open Foundation Models by 01.AI |
| MMLU | 57.87accuracy (%) | unverified· optimizedT1 | 2024-03-01Yi: Open Foundation Models by 01.AI |
| Winogrande | 46accuracy (%) | unverified· optimizedT1 | 2024-03-01Yi: Open Foundation Models by 01.AI |
| ARC AI2 | 40.8accuracy (%) | unverified· optimizedT1 | 2024-03-01Yi: Open Foundation Models by 01.AI |
Claims drawn from cited facts, not live model generation.
Released in 2024 by 01.AI, this model is a mid-size 9-billion-parameter design. That scale implies a pragmatic balance: it favors efficiency and lower cost rather than the extra headroom of a larger frontier-scale model.
3 cited facts
This model is scored on four tracked benchmarks and currently holds no top score in any of them. Because outside evaluation has neither confirmed nor disputed these numbers, the record is unverified and should be read with that uncertainty in mind.
3 cited facts
With an average SOTA gap of 34.2 points, this model sits well behind the leaders on its benchmarks, a wide gap that places it clearly outside the frontier rather than merely competitive. It holds the top score on none of the tracked benchmarks, so there is no category leadership to point to yet.
2 cited facts
2 facts cross-checked across data sources: 2 single-source.
Source facts, citations, and refresh stamp for this record.
Sources: curated_param_counts, epoch_benchmarks
Yi-9B is an AI model developed by 01.AI, released 2024-03-01. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.
Yi-9B has recorded scores on 4 benchmarks, each shown with its evidence status.
Yi-9B has recorded scores on 4 benchmarks — Winogrande, MMLU, ARC AI2, HellaSwag. The full table above shows each score with its evidence status.
0 of 4 recorded scores (0%) are independently reproduced rather than self-reported by the lab.
Yi-9B has 4 tracked claims: 4 unverified.