Skip to content

docs: explain evaluation and prompt feedback settings - #592

Merged
codelion merged 1 commit into
algorithmicsuperintelligence:mainfrom
Yushu-Li:docs/evaluator-prompt-config-284
Nov 7, 2026
Merged

codelion merged 1 commit into
algorithmicsuperintelligence:mainfrom
Yushu-Li:docs/evaluator-prompt-config-284

Conversation

@Yushu-Li

@Yushu-Li Yushu-Li commented Nov 7, 2026

Copy link
Copy Markdown
Contributor

Summary

I added a configuration guide for the cascade, LLM judging, and prompt-template variation settings discussed in #284. The guide distinguishes the scoring paths from prompt wording and shows how to return a diagnostic as an evaluator artifact.

Validation

  • Ran both documented Python blocks verbatim. The invalid candidate stops at stage 1 and its diagnostic appears in the rendered prompt; a valid candidate reaches stage 2.
  • Ran 18 existing cascade tests and 15 prompt sampler tests; all passed. git diff --check passed.
  • No provider calls or production code changes.

I used Codex to help write and verify the example.

@codelion
codelion merged commit 56b2fcf into algorithmicsuperintelligence:main Nov 7, 2026
4 checks passed
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