Violet_Twilight-v0.2: benchmark results
Epiculous's SLERP merge of its Azure_Dusk and Crimson_Dawn v0.2 tunes, a roleplay-leaning Mistral Nemo 12B model. Provider: Other. Released 2024-09-12. Access: Open.
Unified ELO 1504 ± 1, rank #675 of 1392 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - MuSR | 13.61 | Score | 74.8 |
| UGI - Writing | 34.55 | Writing Score | 53.1 |
| Open LLM Leaderboard - IFEval | 45.32 | Score | 50.1 |
| Open LLM Leaderboard - MMLU-Pro | 23.45 | Score | 40.5 |
| UGI - Natural Intelligence | 18.96 | NatInt Score | 36.6 |
| Open LLM Leaderboard - BBH | 23.94 | Score | 36.2 |
| UGI - Willingness (W/10) | 4 | W/10 Score | 30.4 |
| UGI Leaderboard | 25.86 | UGI Score | 21.1 |
| Open LLM Leaderboard - GPQA | 2.13 | Score | 20.3 |
| Open LLM Leaderboard - MATH Level 5 | 2.87 | Score | 18.2 |
Interactive version: theaggregate.ai/model?slug=violet-twilight-v0-2 · How It Works · Data refreshed daily, snapshot 2026-09-05.