RLTasksby ulam.ai
Models / Microsoft

Microsoft / MODEL EVIDENCE

Phi-4

9 benchmarks · 11 observations · 9 sources · released 2024-12-12

Scores at a glance

9 benchmarks · large dot: Phi-4; small dots: closest models where reported

BALROG ↗11.6%
3.768.3

Phi-4 · Balrog · Average progress · % ↗

UGPhysics ↗37.16%
5.3657.16

Phi-4 · UGPhysics · English · overall accuracy · original evaluation ↗

2 more reported results

33.44% · Phi-4 · UGPhysics · Chinese · overall accuracy · original evaluation · UGPhysics: A Comprehensive Benchmark for Undergraduate Physics Reasoning with Large Language Models ↗

35.3% · Phi-4 · UGPhysics · Language average · overall accuracy · original evaluation · UGPhysics: A Comprehensive Benchmark for Undergraduate Physics Reasoning with Large Language Models ↗

Closest models by score

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

0.9 percentile-point gap

3 of 3 shared benchmark comparisons

1.2 percentile-point gap

3 of 3 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. Phi-4 Technical Report ↗

    Marah Abdin, Jyoti Aneja, Harkirat Behl, Sébastien Bubeck, Ronen Eldan, Suriya Gunasekar et al. · 2024

    All 27 authors

    Marah Abdin, Jyoti Aneja, Harkirat Behl, Sébastien Bubeck, Ronen Eldan, Suriya Gunasekar, Michael Harrison, Russell J. Hewett, Mojan Javaheripi, Piero Kauffmann, James R. Lee, Yin Tat Lee, Yuanzhi Li, Weishung Liu, Caio C. T. Mendes, Anh Nguyen, Eric Price, Gustavo de Rosa, Olli Saarikivi, Adil Salim, Shital Shah, Xin Wang, Rachel Ward, Yue Wu, Dingli Yu, Cyril Zhang, Yi Zhang

  2. UGPhysics: A Comprehensive Benchmark for Undergraduate Physics Reasoning with Large Language Models — published result table ↗

    Xin Xu, Qiyun Xu, Tong Xiao, Tianhao Chen, Yuchen Yan, Jiaxin Zhang et al. · 2025

    All 9 authors

    Xin Xu, Qiyun Xu, Tong Xiao, Tianhao Chen, Yuchen Yan, Jiaxin Zhang, Shizhe Diao, Can Yang, Yang Wang

  3. BALROG: Benchmarking Agentic LLM and VLM Reasoning On Games ↗

    Paglieri, Davide, Cupiał, Bartłomiej, Coward, Samuel, Piterbarg, Ulyana, Wolczyk, Maciej, Khan, Akbir et al. · 2024

    All 13 authors

    Paglieri, Davide, Cupiał, Bartłomiej, Coward, Samuel, Piterbarg, Ulyana, Wolczyk, Maciej, Khan, Akbir, Pignatelli, Eduardo, Kuciński, Łukasz, Pinto, Lerrel, Fergus, Rob, Foerster, Jakob Nicolaus, Parker-Holder, Jack, Rocktäschel, Tim

  4. GPQA Diamond ↗

    Epoch AI

  5. MATH Level 5 ↗

    Epoch AI

  6. OTIS Mock AIME 2024-2025 ↗

    Epoch AI

  7. AI benchmarks & capabilities ↗

    Epoch AI

  8. Writing benchmark — incorporating mandatory story elements ↗

    lechmazur

  9. LiveBench — model evaluation leaderboard ↗

    LiveBench Team

Names used in source tables

Phi-4