Le_Triomphant-ECE-TW3: benchmark results
Mergekit merge of Smaug-72B and MultiVerse_70B by paloalma, a student project from France's ECE engineering school with TW3 Partners. Provider: ECE & TW3 Partners. Released 2024-04-01. Access: Open.
Aggregate Rating 1566 ± 22, rank #559 of 1656 rated models, from 20 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - ARC Challenge | 78.5 | Normalized accuracy (%) (25-shot) | 99.8 |
| Open LLM Leaderboard v1 - HellaSwag | 90.3 | Normalized accuracy (%) (10-shot) | 99.7 |
| Open LLM Leaderboard v1 - GSM8K | 79.83 | Accuracy (%) (5-shot) | 99.6 |
| Open LLM Leaderboard v1 - MMLU | 77.81 | Accuracy (%) (5-shot) | 99 |
| Open LLM Leaderboard v1 - WinoGrande | 85.56 | Accuracy (%) (5-shot) | 98.6 |
| Open Chinese LLM - ARC Challenge | 69.2 | Accuracy (%) | 98.5 |
| Open Chinese LLM - GSM8K | 71.57 | Accuracy (%) | 98.5 |
| Open Chinese LLM - HellaSwag | 76.54 | Accuracy (%) | 98.2 |
| Open Chinese LLM - C-Eval Semantic | 91.1 | Accuracy (%) | 97.9 |
| Open Chinese LLM - TruthfulQA MC | 66.42 | Accuracy (%) | 97.9 |
| Open Chinese LLM Leaderboard | 72 | Official Nonzero-Task Mean (%) | 97.9 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 75.84 | MC2 (%) (0-shot) | 95.8 |
Interactive version: theaggregate.ai/model?slug=le-triomphant-ece-tw3 · How It Works · Data refreshed daily, snapshot 2026-10-20.