docs: GLM-5.3-Flash publication note + checkpoint-sentinel measurement lesson - #101
Merged
Conversation
…t lesson The GLM-5.3-Flash row's sibling finding joins the measurement- discipline paragraph: resumable LLM-eval checkpoints silently resume error sentinels (140 exhausted-retry empties read as 15.96 DER until re-fetched; recorded 8.5721/8.7978). PUBLICATION-NOTES gains the frontier-regression finding with the two live-discovered protocol facts (thinking cannot be disabled; efforts exactly low/high/max).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Companion to PR #100 (leaderboard row): the PUBLICATION-NOTES finding for the frontier-regression result, and the sibling measurement lesson — resumable LLM-eval checkpoints silently resume error sentinels (140 exhausted-retry empties read as 15.96 DER until detected; the recorded numbers are 8.5721 raw / 8.7978 zero-skip). Placed next to the subset-inflation instances in the paper's measurement-discipline paragraph. The resume path now drops empty rows by default (rababa PR #65).