Add mDeBERTa-RAID-v1 submission - #206
Open
Lendarixon wants to merge 1 commit into
Open
Lendarixon wants to merge 1 commit into
Lendarixon wants to merge 1 commit into
Conversation
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Lendarixon
requested a deployment
to
raid-main
September 26, 2026 18:24 — with 
GitHub Actions
Waiting
GitHub Actions
Waiting
This branch is waiting to be deployed
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds
leaderboard/submissions/mDeBERTa-RAID-v1/(predictions.json for all 672,000 test ids + metadata.json).Detector: mDeBERTa-v3-base (multilingual EN/RU/PL AI-text classifier), fine-tuned on a 362k-text sample of the RAID train split (all 8 domains, all 11 generators, all 12 attacks; human texts of every attack labelled human) plus a small replay of its original EN/RU/PL training data. Before scoring, input text is normalized against character-level attacks (homoglyphs, zero-width characters, unusual whitespace). Long texts are scored in 256-token windows covering the whole document; the score is the mean logit margin (AI minus human), z-normalized.
Local validation: 15% of RAID train source documents were held out (no text of those documents was used in training), scored with
raid.evaluate.run_evaluation(per-domain thresholds, FPR 5%): 96.1% accuracy with adversarial attacks, 97.2% without.🤖 Generated with Claude Code