pythia-410m-deduped: benchmark results

Provider: EleutherAI. Access: Open.

Aggregate Rating 1120 ± 17, rank #10509 of 10916 rated models, from 73 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
MMLU-by-task - High School Statistics47.22Accuracy (%)83.3
MMLU-by-task - Machine Learning33.93Accuracy (%)57.4
MMLU-by-task - Formal Logic30.95Accuracy (%)52.2
MMLU-by-task - High School Mathematics26.67Accuracy (%)45.9
MMLU-by-task - College Chemistry31Accuracy (%)43.8
MMLU-by-task - College Physics23.53Accuracy (%)41.6
MMLU-by-task - College Biology32.64Accuracy (%)35.4
MMLU-by-task - Security Studies35.1Accuracy (%)32.4
MMLU-by-task - TruthfulQA MC240.95Accuracy (%)31.1
MMLU-by-task - Elementary Mathematics26.19Accuracy (%)30
MMLU-by-task - Medical Genetics31Accuracy (%)29.8
MMLU-by-task - Jurisprudence29.63Accuracy (%)29.6

Interactive version: theaggregate.ai/model?slug=pythia-410m-deduped · How It Works · Data refreshed daily, snapshot 2026-11-08.