granite-4.0-1B-base: benchmark results

Provider: IBM. Access: Open.

Aggregate Rating 1361 ± 34, rank #7383 of 10552 rated models, from 26 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Lara Mini Base Report (base-model protocol) - CRUXEval-I (2-shot)39.6pass@1, direct prediction, official answer parser (%)86.1
Lara Mini Base Report (base-model protocol) - CRUXEval-O (2-shot)36.1pass@1, direct prediction, official answer parser (%)70.4
Lara Mini Base Report (base-model protocol) - MBPP (3-shot)46.4pass@1 (execution) (%)61.1
Lara Mini Base Report (base-model protocol) - GSM8K (5-shot)59.4Exact match (strict extraction: first "#### N", else first \59.3
Lara Mini Base Report (base-model protocol) - MATH (4-shot)35.5Math-Verify equivalence, Minerva prompts, 5,000 problems ove55.6
Lara Mini Base Report (base-model protocol) - MBPP+ (3-shot)49.7pass@1; must pass MBPP's own and the EvalPlus tests (%)55.6
Lara Mini Base Report (base-model protocol) - MGSM (8 (CoT)-shot)35Exact match on each language's native CoT answer phrase, ave51.9
Lara Mini Base Report (base-model protocol) - WinoGrande (5-shot)67.6Accuracy (%), English benchmark50.9
Lara Mini Base Report (base-model protocol) - COPA (0-shot)82.6Accuracy (%), English benchmark48.1
Lara Mini Base Report (base-model protocol) - HellaSwag (10-shot)70.2Accuracy normalized by answer length (%), English benchmark48.1
Lara Mini Base Report (base-model protocol) - MMLU (5-shot)57.2Accuracy (%), English benchmark48.1
Lara Mini Base Report (base-model protocol) - Belebele, All Languages (0-shot)32.2Accuracy (%) averaged over all languages/variants of the ben47.2

Interactive version: theaggregate.ai/model?slug=granite-4-0-1b-base · How It Works · Data refreshed daily, snapshot 2026-11-16.