RLTasksby ulam.ai
Models / Anthropic

Anthropic / MODEL EVIDENCE

Claude 3.5 Haiku

5 benchmarks · 5 observations · 5 sources · released 2024-10-22

Scores at a glance

5 benchmarks · large dot: Claude 3.5 Haiku; small dots: closest models where reported

Fortress ↗30.41%
5.7874.39

Claude 3.5 Haiku · Scale Labs · Risk elicitation score ↗

FACTS Grounding ↗74.2%
61.783.6

Claude 3.5 Haiku · FACTS Grounding · Final factuality score · judge average · eligibility-filtered evaluation ↗

Closest models by score

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

15.4 percentile-point gap

3 of 4 shared benchmark comparisons

18.7 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. HARP: A challenging human-annotated math reasoning benchmark — published result table ↗

    Yue, Albert S., Madaan, Lovish, Moskovitz, Ted, Strouse, DJ, Singh, Aaditya K. · 2024

  2. The FACTS Grounding Leaderboard: Benchmarking LLMs' Ability to Ground Responses to Long-Form Input — published result table ↗

    Alon Jacovi, Andrew Wang, Chris Alberti, Connie Tao, Jon Lipovetz, Kate Olszewska et al. · 2025

    All 26 authors

    Alon Jacovi, Andrew Wang, Chris Alberti, Connie Tao, Jon Lipovetz, Kate Olszewska, Lukas Haas, Michelle Liu, Nate Keating, Adam Bloniarz, Carl Saroufim, Corey Fry, Dror Marcus, Doron Kukliansky, Gaurav Singh Tomar, James Swirhun, Jinwei Xing, Lily Wang, Madhu Gurumurthy, Michael Aaron, Moran Ambar, Rachana Fellinger, Rui Wang, Zizhao Zhang, Sasha Goldshtein, Dipanjan Das

  3. MedXpertQA: Benchmarking Expert-Level Medical Reasoning and Understanding — published result table ↗

    Yuxin Zuo, Shang Qu, Yifei Li, Zhangren Chen, Xuekai Zhu, Ermo Hua et al. · 2025

    All 9 authors

    Yuxin Zuo, Shang Qu, Yifei Li, Zhangren Chen, Xuekai Zhu, Ermo Hua, Kaiyan Zhang, Ning Ding, Bowen Zhou

  4. AI benchmarks & capabilities ↗

    Epoch AI

  5. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

Names used in source tables

Claude 3.5 Haiku · Claude-3.5-Haiku