gpt-neo-1.3B: benchmark results
EleutherAI's GPT-Neo 1.3B, an open replication of the GPT-3 architecture trained on the Pile, the mid-sized model in EleutherAI's original GPT-Neo release. Provider: EleutherAI. Released 2021-03-21. Access: Open.
Aggregate Rating 1123 ± 15, rank #10470 of 10916 rated models, from 80 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MMLU-by-task - High School Statistics | 40.74 | Accuracy (%) | 57.1 |
| MMLU-by-task - Abstract Algebra | 28 | Accuracy (%) | 39.6 |
| MMLU-by-task - Machine Learning | 29.46 | Accuracy (%) | 35.2 |
| MMLU-by-task - College Computer Science | 32 | Accuracy (%) | 32.6 |
| Open LLM Leaderboard - MuSR | 38.17 | Score | 30.2 |
| MMLU-by-task - Medical Genetics | 31 | Accuracy (%) | 29.8 |
| MMLU-by-task - Formal Logic | 25.4 | Accuracy (%) | 28.9 |
| MMLU-by-task - Electrical Engineering | 28.28 | Accuracy (%) | 28.1 |
| MMLU-by-task - High School Macroeconomics | 31.28 | Accuracy (%) | 28.1 |
| MMLU-by-task - Marketing | 29.49 | Accuracy (%) | 28 |
| MMLU-by-task - World Religions | 32.16 | Accuracy (%) | 28 |
| MMLU-by-task - High School World History | 28.27 | Accuracy (%) | 27.9 |
Interactive version: theaggregate.ai/model?slug=gpt-neo-1-3b · How It Works · Data refreshed daily, snapshot 2026-11-08.