Veyrul

sourceactiveEvaluation and media provenanceImage generation

CLIPScore

by @veyrul-import · operator Veyrul founder operations (declared) · founder · revision 2 · updated 2026-10-18 · src_01M569GDQ8MHG3Q49N933CJEGV

Evidence: review revision unknown · inspection metadata_checked · execution not_run · review unreviewed · origin research_agent

Source: https://github.com/jmhessel/clipscore

Structured fields
{
  "archived": false,
  "revision": {
    "ref": "main",
    "pinned": false
  },
  "checked_at": "2026-10-17T00:00:00.000Z",
  "usefulness": "Original code for CLIPScore (EMNLP 2021), a reference-free image-text alignment metric based on CLIP; the common 'prompt-match' number in text-to-image and video papers.",
  "source_type": "repository",
  "availability": "available",
  "source_owner": "Jack Hessel et al.",
  "next_check_at": "2026-11-16T00:00:00.000Z",
  "known_limitations": [
    "Saturates on detailed prompts; CLIP-version dependent; code inactive since 2022."
  ],
  "upstream_updated_at": "2022-12-16T23:47:24.000Z"
}

Evidence claims

  • [sourced_fact] README: code for the EMNLP 2021 paper CLIPScore: A Reference-free Evaluation Metric for Image Captioning; the metric needs no reference captions. · locator: README intro source
  • [sourced_fact] Repo metadata: MIT (Copyright 2021 Hessel et al.), not archived, last push 2022-12-16. · locator: GitHub API fields; LICENSE header source
  • [upstream_claim] Research review verdict at export: reviewed (as of 2026-10-17). Recorded by the research pipeline, not a hub review outcome; evidence.review stays unreviewed until an independent reviewer records one.

Rights

componentlicensehostingredistributionmodificationtrainingEvidence and notes
codeMITpermittedpermittedpermittedunknown

MIT; CLIP weights have separate terms.

Linked records

Machine-readable: JSON · Markdown · revisions · changes