fix: improve sentiment scoring accuracy#50
Merged
Conversation
Address systematic bearish bias identified by audit (r=0.43, 50% agreement): - Fix duplicate segments bug in backfill: store only selftext in content field instead of pre-concatenating comments (which were scored twice) - Add paragraph deduplication in _extract_user_info() for historical data - Reduce title weight from 0.4 to 0.2 so comment consensus drives scores - Reduce tanh amplification from 5x to 3x for more gradual score curve - Add 33 neutral anchor phrases for Q&A, tool comparisons, and token burns that were being mis-classified as bearish Eval on 154 labeled posts: MAE 0.131→0.086 (-34%), categorical agreement 46.8%→51.3%, bearish bias -0.050→-0.029. Pearson r ~0.44 (unchanged — limited by cosine-similarity approach ceiling). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Documents the 154-post labeled evaluation, metrics (Pearson r, MAE, categorical agreement, bias), root causes found, fixes applied, and before/after results. Includes reproduction steps and next-steps analysis for the scoring approach ceiling. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
_extract_user_info()Evaluation (154 human-labeled posts)
Full methodology and analysis in
docs/scoring_accuracy_audit.md.Test plan
uv run pytest tests/ -x)🤖 Generated with Claude Code