Napoleon_24B_V0.2 — benchmark results

baconnier's DARE-TIES re-merge blending Dolphin 3.0 Mistral 24B (60%) back with his earlier Napoleon_24B_V0.0 fine-tune (40%). Provider: Other. Released 2025-02-17. Access: Open.

Unified ELO 1420 ± 46, rank #1132 of 1776 rated models, from 10 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open LLM Leaderboard - GPQA11.74Score83.7
Open LLM Leaderboard - BBH41.21Score80.4
Open LLM Leaderboard - MMLU-Pro37.3Score79.3
Open LLM Leaderboard - MuSR14.18Score77.6
Open LLM Leaderboard - MATH Level 514.35Score60.8
French LLM Leaderboard - Average26.09Average Score (%)25.9
French LLM Leaderboard - BAC FR19.92Score (%)25.9
Open LLM Leaderboard - IFEval25.27Score21.3
French LLM Leaderboard - IFEval FR32.73Score (%)20.4
French LLM Leaderboard - GPQA FR25.64Score (%)16.7

Interactive version: theaggregate.ai/model?slug=napoleon-24b-v0-2 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.