Pıer
TidesCurrentsHarbor LightsLabBottlesAshore
Pıer

Navigation

  • Tides
  • Ashore
  • Harbor Lights
  • Agent Access
  • Changelog
  • Bottles
  • Now
  • Feedback

External links

GitHubCloudborne ↗

© 2026 Pier.

WatchingResearchWatching0 independent reports0

GAMUT: Two-Level Meta-Rubrics for Evaluating Factual Completeness in Open-Ended Generation

First seen · 7/22/2026, 12:00 PMLatest activity · 7/22/2026, 12:00 PM

The paper introduces a two-level meta-rubric framework for evaluating factual completeness, complementing precision-focused methods that mainly detect false claims. A structured rubric captures the organization, importance, open-ended sets, ordered processes, and relationships required in a complete answer, then compiles these requirements into machine-gradable binary checks. GAMUT contains 1,813 questions grounded in real wearable imagery across 10 domains, with an additional text-only variant. Evaluation of 14 frontier and open-weight models reports a best score of 58.7% for Gemini 3.1 Pro, while claiming strong discrimination and robustness across judges.

Event heat · last 24 hours

No heat snapshots are available in the last 24 hours.

No heat snapshots are available in the last 24 hours.

Reporting Timeline

  1. AggregatorHuggingFace Daily Papers7/22, 12:00 PMnot independentRepresentative
    GAMUT: Two-Level Meta-Rubrics for Evaluating Factual Completeness in Open-Ended Generation