Pallas-0.2 [bfloat16] [rev: 2a367db35e91]: benchmark results

Provider: Mihaiii. Access: Open.

Aggregate Rating 1474 ± 37, rank #2758 of 10916 rated models, from 6 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open LLM Leaderboard v1 - MMLU75.53Accuracy (%) (5-shot)96.9
Open LLM Leaderboard v1 - WinoGrande81.61Accuracy (%) (5-shot)80.1
Open LLM Leaderboard v1 - GSM8K62.62Accuracy (%) (5-shot)75.9
Open LLM Leaderboard v1 - ARC Challenge64.59Normalized accuracy (%) (25-shot)67.3
Open LLM Leaderboard v1 - TruthfulQA MC255.29MC2 (%) (0-shot)64.7
Open LLM Leaderboard v1 - HellaSwag83.44Normalized accuracy (%) (10-shot)58.6

Interactive version: theaggregate.ai/model?slug=pallas-0-2-bfloat16-rev-2a367db35e91 · How It Works · Data refreshed daily, snapshot 2026-11-08.