galactica-1.3B: benchmark results
Meta AI Papers with Code Galactica 1.3B is a scientific decoder-only language model trained on a 106B-token scientific corpus, including papers, textbooks and structured scientific data. This is the original base release under CC BY-NC 4.0, separate from OPT and instruction-tuned variants. Provider: Other. Access: Open.
Aggregate Rating 1143 ± 18, rank #10064 of 10916 rated models, from 67 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MMLU-by-task - High School Physics | 32.45 | Accuracy (%) | 66 |
| MMLU-by-task - Global Facts | 35 | Accuracy (%) | 64.2 |
| MMLU-by-task - High School Mathematics | 27.04 | Accuracy (%) | 50.6 |
| MMLU-by-task - Machine Learning | 30.36 | Accuracy (%) | 39 |
| MMLU-by-task - Moral Scenarios | 24.58 | Accuracy (%) | 36.2 |
| MMLU-by-task - High School Chemistry | 29.06 | Accuracy (%) | 36.1 |
| MMLU-by-task - Professional Psychology | 30.23 | Accuracy (%) | 34.6 |
| MMLU-by-task - Computer Security | 37 | Accuracy (%) | 33.9 |
| MMLU-by-task - Anatomy | 34.07 | Accuracy (%) | 33.8 |
| MMLU-by-task - High School European History | 30.3 | Accuracy (%) | 33.6 |
| MMLU-by-task - TruthfulQA MC2 | 41.4 | Accuracy (%) | 33.3 |
| MMLU-by-task - Logical Fallacies | 30.67 | Accuracy (%) | 33.2 |
Interactive version: theaggregate.ai/model?slug=galactica-1-3b · How It Works · Data refreshed daily, snapshot 2026-11-08.