RWKV-4-PilePlus-169M-20230520-done-ctx4096: benchmark results

Provider: RWKV. Access: Open.

Aggregate Rating 1117 ± 16, rank #10542 of 10916 rated models, from 68 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
MMLU-by-task - Machine Learning30.36Accuracy (%)39
MMLU-by-task - TruthfulQA MC242.29Accuracy (%)38.4
MMLU-by-task - Formal Logic26.19Accuracy (%)32.4
MMLU-by-task - Computer Security32Accuracy (%)29.1
MMLU-by-task - College Biology27.78Accuracy (%)28.9
MMLU-by-task - Virology31.33Accuracy (%)28.1
MMLU-by-task - Business Ethics30Accuracy (%)26.7
MMLU-by-task - College Physics21.57Accuracy (%)25.7
MMLU-by-task - Medical Genetics30Accuracy (%)25.6
MMLU-by-task - World Religions30.41Accuracy (%)22.4
MMLU-by-task - Human Sexuality25.95Accuracy (%)20.9
MMLU-by-task - Nutrition26.47Accuracy (%)20.6

Interactive version: theaggregate.ai/model?slug=rwkv-4-pileplus-169m-20230520-done-ctx4096 · How It Works · Data refreshed daily, snapshot 2026-11-08.