Midm-2.0-Base-Instruct: benchmark results

Provider: Other.

Unified ELO 1498 ± 31, rank #765 of 1607 rated models, from 39 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Polar (US) - Sociocultural Axis84.25Sociocultural-axis ICAT (0-100), the StereoSet idealized con75.7
Horangi 4 - KorQuAD 1.085.15F1 (x100)73.1
Horangi 4 - KoBBQ92Accuracy (%)72.6
Polar (South Korea) - Economic Axis87.24Economic-axis ICAT (0-100), the StereoSet idealized context 67.6
Polar (US)76.12Total ICAT (0-100), the StereoSet idealized context associat64.9
Polar (South Korea)88.17Total ICAT (0-100), the StereoSet idealized context associat59.5
Polar (South Korea) - Sociocultural Axis89.09Sociocultural-axis ICAT (0-100), the StereoSet idealized con51.4
Horangi 4 - HAE-RAE Bench (without Reading Comprehension)81.82Accuracy (%)34.6
Polar (US) - Economic Axis67.99Economic-axis ICAT (0-100), the StereoSet idealized context 32.4
Horangi 4 - Ko-HalluLens (Nonexistent Entities)64Refusal rate (%)32.2
Horangi 4 - Ko-HLE10Accuracy (%)29.3
Horangi 4 - GLP - General Knowledge73.91Score (%)28.8

Interactive version: theaggregate.ai/model?slug=midm-2-0-base-instruct · How It Works · Data refreshed daily, snapshot 2026-09-29.