SOLAR-10.7B-Instruct-v1.0-uncensored — benchmark results
w4r10ck's uncensored variant of Upstage's SOLAR 10.7B Instruct, LoRA/DPO-trained on the unalignment toxic-dpo-v0.1 dataset to relax refusals (December 2023). Provider: Upstage. Released 2023-12-14. Access: Open.
Unified ELO 1458 ± 36, rank #961 of 1776 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - MuSR | 18.49 | Score | 91.1 |
| UGI - Willingness (W/10) | 7.5 | W/10 Score | 77.6 |
| Open LLM Leaderboard - BBH | 33.86 | Score | 66.6 |
| UGI Leaderboard | 35.12 | UGI Score | 52.3 |
| Open LLM Leaderboard - GPQA | 5.93 | Score | 49.7 |
| Open LLM Leaderboard - MMLU-Pro | 26.04 | Score | 47.3 |
| Open LLM Leaderboard - IFEval | 38.84 | Score | 38.5 |
| Open LLM Leaderboard - MATH Level 5 | 6.57 | Score | 35.6 |
| UGI - Writing | 21.9 | Writing Score | 16.9 |
| UGI - Natural Intelligence | 12.39 | NatInt Score | 10.5 |
Interactive version: theaggregate.ai/model?slug=solar-10-7b-instruct-v1-0-uncensored · How the rankings work · Data refreshed daily, snapshot 2026-07-22.