Llama-3SOME-8B-v2 — benchmark results

TheDrummer's (BeaverAI) NSFW roleplay fine-tune of Llama 3 8B, the v2 of his 3SOME series tuned for adult creative writing. Provider: Meta. Released 2024-06-21. Access: Open.

Unified ELO 1438 ± 21, rank #1054 of 1776 rated models, from 10 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open LLM Leaderboard - MMLU-Pro30.59Score62.6
Open LLM Leaderboard - BBH31.7Score58.2
Open LLM Leaderboard - GPQA6.94Score58.2
Open LLM Leaderboard - IFEval45.08Score49.6
Open LLM Leaderboard - MATH Level 59.37Score45.3
UGI - Writing29.48Writing Score35.3
Open LLM Leaderboard - MuSR7.21Score33.8
UGI - Willingness (W/10)3.8W/10 Score27.8
UGI - Natural Intelligence14.31NatInt Score16.1
UGI Leaderboard19.18UGI Score10.5

Interactive version: theaggregate.ai/model?slug=llama-3some-8b-v2 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.