DialoGPT-small: benchmark results
Microsoft's DialoGPT dialogue-response model for multi-turn conversations, trained on 147 million multi-turn dialogues from Reddit discussion threads. This release retains its distinct size variant; the dialogue count is not a training-token count. Provider: Microsoft. Access: Open.
Aggregate Rating 1127 ± 18, rank #10401 of 10916 rated models, from 68 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MMLU-by-task - High School Statistics | 43.52 | Accuracy (%) | 68.1 |
| MMLU-by-task - TruthfulQA MC2 | 47.49 | Accuracy (%) | 62.2 |
| MMLU-by-task - Formal Logic | 31.75 | Accuracy (%) | 56 |
| MMLU-by-task - Professional Medicine | 44.85 | Accuracy (%) | 45.7 |
| MMLU-by-task - College Medicine | 36.42 | Accuracy (%) | 43.2 |
| MMLU-by-task - High School Macroeconomics | 36.41 | Accuracy (%) | 39.5 |
| MMLU-by-task - High School Microeconomics | 34.87 | Accuracy (%) | 39.3 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 47.49 | MC2 (%) (0-shot) | 37.9 |
| MMLU-by-task - High School Government And Politics | 36.79 | Accuracy (%) | 33.5 |
| MMLU-by-task - High School Geography | 35.35 | Accuracy (%) | 32.2 |
| MMLU-by-task - High School Biology | 30.97 | Accuracy (%) | 31.2 |
| MMLU-by-task - Security Studies | 32.65 | Accuracy (%) | 29.6 |
Interactive version: theaggregate.ai/model?slug=dialogpt-small · How It Works · Data refreshed daily, snapshot 2026-11-08.