LMSYS·released 2023-04-121 source
vicuna-13b-v1.1 benchmark scores: 7 benchmarks tracked. 0% of its scores are independently reproduced — the rest are self-reported or unverified.
Leads on 0 of 7 benchmarks·best result 54.8 on PIQA (self-reported)·0 of 7 independently reproduced
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| PIQA | 54.8accuracy (%) | self-reported· optimizedT1 | 2023-04-12Textbooks Are All You Need II: phi-1.5 technical report |
| HellaSwag | 43.73accuracy (%) | self-reported· optimizedT1 | 2023-04-12Textbooks Are All You Need II: phi-1.5 technical report |
| Winogrande | 41.6accuracy (%) | self-reported· optimizedT1 | 2023-04-12Textbooks Are All You Need II: phi-1.5 technical report |
| GSM8K | 28.13accuracy (%) | unverified· optimizedT1 | 2023-04-12Baichuan 2: Open Large-scale Language Models |
| ARC AI2 | 24.27accuracy (%) | self-reported· optimizedT1 | 2023-04-12Textbooks Are All You Need II: phi-1.5 technical report |
| BBH | 24.05accuracy (%) | unverified· optimizedT1 | 2023-04-12Baichuan 2: Open Large-scale Language Models |
| OpenBookQA | 10.67accuracy (%) | self-reported· optimizedT1 | 2023-04-12Textbooks Are All You Need II: phi-1.5 technical report |
Claims drawn from cited facts, not live model generation.
Developed by LMSYS, this model is a 13-billion-parameter mid-size release from 2023. That scale implies a practical middle ground: enough headroom for many tasks while keeping inference more efficient and cheaper than frontier-scale models.
4 cited facts
This model is tracked across 7 benchmarks and currently holds no top score on any of them. Because these results are self-reported and not yet independently confirmed, the record should be read as vendor claims rather than verified outcomes.
3 cited facts
With an average SOTA gap of 54.19 points, this model trails the leader by a wide margin on the disclosed harness, placing it well behind the front of the pack. It holds the top score on none of the tracked benchmarks, meaning it has not achieved category leadership on any of them.
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
vicuna-13b-v1.1 is an AI model developed by LMSYS, released 2023-04-12. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.
vicuna-13b-v1.1 has recorded scores on 7 benchmarks, each shown with its evidence status.
vicuna-13b-v1.1 has recorded scores on 7 benchmarks — PIQA, OpenBookQA, Winogrande, ARC AI2, BBH, GSM8K, and 1 more. The full table above shows each score with its evidence status.
0 of 7 recorded scores (0%) are independently reproduced rather than self-reported by the lab.
vicuna-13b-v1.1 has 7 tracked claims: 5 self-reported, 2 unverified.