Skip to content

docs(evaluation): Judge models concept#726

Merged
khushalsonawat merged 4 commits into
docs/evaluation-revampfrom
docs/eval-judge-models
Jul 10, 2026
Merged

docs(evaluation): Judge models concept#726
khushalsonawat merged 4 commits into
docs/evaluation-revampfrom
docs/eval-judge-models

Conversation

@YashMohan23

@YashMohan23 YashMohan23 commented Jul 9, 2026

Copy link
Copy Markdown

The Judge models concept page (/docs/evaluation/concepts/judge-models).

  • What a judge is: reads and scores a response, never edits it
  • The Turing/Protect model table stays on this page
  • Bring your own judge, and modality (image/audio need a multimodal judge)
  • How to choose

Also fixes a factual error from the old page: it listed turing_small for audio, but turing_small is text and image only.

What a judge is (reads and scores, never edits the output), the Turing/Protect
model table on this page, bring-your-own judge, modality (image/audio need a
multimodal judge), and how to choose. Fixed the audio-modality error (turing_small
is text+image only). Keep-exploring cards avoid the prev/next neighbors. No About,
minimal frontmatter, no em-dashes, all links resolve.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
@YashMohan23 YashMohan23 marked this pull request as ready for review July 9, 2026 09:24
Yash Mohan and others added 3 commits July 9, 2026 17:06
…scoring diagram

Open directly at 'What a judge does' with no lead paragraph, add a diagram of how a
judge scores (response and criteria -> judge -> result and optional reason), link
first mentions (template, eval types), soften reason to optional, and shorten the description.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
"Judge models" read as the LLM-as-Judge eval type (Rishav's feedback),
so the concept is now Evaluator model: page moved to
concepts/evaluator-models, term renamed across the evaluation tree,
FAQ, nav, and redirect targets; the LLM-as-Judge type name stays.

Also addresses review on the page: stale guides/ links repointed,
modality line made consistent with the table, first mention bolded,
backward card swapped for Guardrails, redundant How to choose section
dropped, and "Bring your own LLM" phrasing everywhere.

Signed-off-by: khushalsonawat <khushalsonawat@gmail.com>
@khushalsonawat khushalsonawat merged commit e5aede7 into docs/evaluation-revamp Jul 10, 2026
@khushalsonawat khushalsonawat deleted the docs/eval-judge-models branch July 10, 2026 07:06
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants