RLTasksby ulam.ai
Models / Google

Google / MODEL EVIDENCE

Gemini 2.0 Flash Exp

4 benchmarks · 6 observations · 6 sources

Scores at a glance

4 benchmarks · large dot: Gemini 2.0 Flash Exp; small dots: closest models where reported

PRMBench ↗66 points
4768.8

Gemini-2.0-flash-exp · PRMBench · Overall PRMScore · original evaluation ↗

Closest models by score

Similar positions on shared benchmark leaderboards. Smaller gaps are closer.

22.1 percentile-point gap

3 of 4 shared benchmark comparisons

o3 →OpenAI
41.0 percentile-point gap

3 of 4 shared benchmark comparisons

How similarity is measured

Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.

Filter comparisons to this model →

Sources

Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.

  1. Aider code editing benchmark ↗

    Paul Gauthier, Aider contributors

  2. Aider refactoring benchmark ↗

    Paul Gauthier, Aider contributors

  3. PRMBench: A Fine-grained and Challenging Benchmark for Process-Level Reward Models — published result table ↗

    Song, Mingyang, Su, Zhaochen, Qu, Xiaoye, Zhou, Jiawei, Cheng, Yu · 2025

  4. MATH 500 — evaluation methodology and results ↗

    Vals AI · 2026

  5. MortgageTax — evaluation methodology and results ↗

    Vals AI · 2026

  6. TaxEval — evaluation methodology and results ↗

    Vals AI · 2026

Names used in source tables

Gemini 2.0 Flash Exp · Gemini-2.0-flash-exp · gemini-2.0-flash-exp