Tiger-Gemma-9B-v3: benchmark results
TheDrummer's v3 decensor of Gemma 2 9B, built on the SPPO variant with a new dataset to cut refusals and verbosity. Provider: TheDrummer. Released 2024-10-04. Access: Open.
Aggregate Rating 1512 ± 47, rank #2162 of 10916 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| UGI - Willingness (W/10) | 8.5 | W/10 Score | 89.6 |
| Open LLM Leaderboard - GPQA | 33.89 | Score | 84.1 |
| Open LLM Leaderboard - IFEval | 68.21 | Score | 80.9 |
| Open LLM Leaderboard - BBH | 58.12 | Score | 79.2 |
| Open LLM Leaderboard - MMLU-Pro | 40.59 | Score | 74.3 |
| Open LLM Leaderboard - MATH Level 5 | 16.24 | Score | 64.1 |
| UGI Leaderboard | 38.75 | UGI Score | 61.4 |
| Open LLM Leaderboard - MuSR | 40.04 | Score | 42.2 |
| UGI - Writing | 28.1 | Writing Score | 30.5 |
| UGI - Natural Intelligence | 16.16 | NatInt Score | 24 |
Interactive version: theaggregate.ai/model?slug=tiger-gemma-9b-v3 · How It Works · Data refreshed daily, snapshot 2026-11-08.