Skip to content

Generate a real pinned TinyDev corpus - #28

Merged
byte271 merged 4 commits into
mainfrom
feat/real-tiny-dev-corpus
Aug 18, 2026
Merged

byte271 merged 4 commits into
mainfrom
feat/real-tiny-dev-corpus

Conversation

@byte271

@byte271 byte271 commented Aug 18, 2026

Copy link
Copy Markdown
Owner

Implements the next M6 prerequisite as an actual generation path rather than another synthetic evidence wrapper.

  • adds a deterministic 104-prompt TinyDev source suite covering the frozen 4-domain × calibration/attack-development matrix
  • records prompt source/license status without inventing a project license
  • generates matched watermarked/unwatermarked pairs with one shared seed per pair under a frozen seed policy
  • retries the whole pair deterministically on early EOS or duplicate text/token outputs
  • records exact generation-token and text-only token tracks in the existing content-addressed CorpusSample schema
  • adds a pinned Hugging Face GPT-2 + SynthID runner at model revision 607a30d783dfa663caf39e06633721c8d4cfcd7e
  • adds an Actions job that produces the full 208-sample real TinyDev artifact
  • adds unit regressions for matrix coverage, exact 64-token pair generation, retry behavior, and fail-closed exhaustion

Scientific note: the frozen spec requires matched seed policy, not necessarily identical actual seeds. This runner intentionally uses the same actual seed within each on/off pair as a stronger controlled-generation choice, without changing the global CorpusManifest contract.

@chatgpt-codex-connector

Copy link
Copy Markdown

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.
To continue using code reviews, you can upgrade your account or add credits to your account and enable them for code reviews in your settings.

@coderabbitai

coderabbitai Bot commented Aug 18, 2026

Copy link
Copy Markdown

Important

Review available on request

  • 🔍 Trigger review

Reviews should be triggered manually for repositories with fewer than 10 stars. Select Trigger review above or comment @coderabbitai review to review the latest changes. For a full review, comment @coderabbitai full review.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: 7b1707b1-afdc-4da1-be50-2cf323c7f268


Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

Comment @coderabbitai help to get the list of available commands.

@byte271
byte271 merged commit 33b4a8c into main Aug 18, 2026
5 checks passed
@github-actions
github-actions Bot deleted the feat/real-tiny-dev-corpus branch August 22, 2026 20:51
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.

1 participant