fix(autodata): fail-loud on empty solver content + tier knobs + the honest live result#42
Merged
Merged
Conversation
…Tokens + tier env knobs The strong reasoning-model solver returned empty visible content at maxTokens=1024 (budget spent on hidden reasoning) and was silently scored 0 — manufacturing a false negative strong/weak gap across every live run. Fix: maxTokens=8000 + throw on empty content (no silent zeros). Make the solver tier an env knob (AUTODATA_WEAK_MODEL/ STRONG_MODEL/CHALLENGER_MODEL/JUDGE_MODEL) + price the wide tier. With the fix the gap is ~0 (not negative): on extractive doc-grounded QA an 8B scores as well as a frontier model — the lever is challenger difficulty, not model tier. See docs/results/autodata-live.md.
tangletools
approved these changes
Jun 25, 2026
tangletools
left a comment
tangletools
left a comment
Contributor
There was a problem hiding this comment.
✅ Auto-approved PR — 2953db28
Blanket team auto-approval is enabled for this reviewer service.
The full PR reviewer audit still runs separately and will publish findings if it detects issues.
tangletools · auto-approval · reason: blanket_auto_approve · 2026-06-25T22:29:59Z
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Autopsy of the Autodata live null: the strong reasoning-model solver returned EMPTY content at maxTokens=1024 (reasoning ate the budget) and was scored 0 → a FALSE negative strong/weak gap across every run. Fix: maxTokens=8000 + fail-loud on empty (no silent zeros), solver tier as env knobs, price table for the wide tier. With the fix the gap is ~0 (not negative): on extractive doc-grounded QA an 8B (llama-3.1-8b) scores as well as a frontier model (gemini-2.5-pro) — so 0 examples discriminate, and the real lever is CHALLENGER difficulty (non-extractive questions), not model tier. Full autopsy + numbers in docs/results/autodata-live.md.