MAmmoTH2-7B-Plus: benchmark results
TIGER Lab (CMU)'s MAmmoTH2-7B-Plus, a Mistral 7B model instruction-tuned on 10 million web-harvested instruction-response pairs (WebInstruct) plus additional public instruction-tuning datasets. Provider: CMU. Released 2024-05-06. Access: Open.
Aggregate Rating 1403 ± 15, rank #1214 of 1656 rated models, from 35 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 71.11 | Accuracy (%) (5-shot) | 95.2 |
| Open Korean LLM - Ko-IFEval | 44.65 | Strict accuracy, prompt/instruction mean (%) | 89 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 39.8 | Exact match, flexible extract (%) | 74.2 |
| Open Chinese LLM - TruthfulQA MC | 56.66 | Accuracy (%) | 71 |
| Open Korean LLM - Ko-EQ-Bench | 24.54 | EQ-Bench score | 70.2 |
| Open Portuguese LLM - FaQuAD NLI | 74.23 | Macro F1 (%) | 70 |
| Open LLM Leaderboard - MATH Level 5 | 18.58 | Score | 68.4 |
| Open LLM Leaderboard - IFEval | 55.75 | Score | 67.6 |
| Open Ko-LLM Leaderboard | 38.34 | Average (%) | 66.3 |
| Open Korean LLM Leaderboard | 38.34 | Average Score (%) | 66.3 |
| Open LLM Leaderboard v1 - ARC Challenge | 63.57 | Normalized accuracy (%) (25-shot) | 63.3 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 54.53 | MC2 (%) (0-shot) | 62.8 |
Interactive version: theaggregate.ai/model?slug=mammoth2-7b-plus · How It Works · Data refreshed daily, snapshot 2026-10-20.