Skip to content

Add DeBERTa-ConPara v2.12, v2.13, v2.14#172

Open
MohamedMady19 wants to merge 2 commits into
liamdugan:mainfrom
MohamedMady19:submission/conpara-v212-v213-v214
Open

Add DeBERTa-ConPara v2.12, v2.13, v2.14#172
MohamedMady19 wants to merge 2 commits into
liamdugan:mainfrom
MohamedMady19:submission/conpara-v212-v213-v214

Conversation

@MohamedMady19

Copy link
Copy Markdown

No description provided.

@MohamedMady19

Copy link
Copy Markdown
Author

Hi @liamdugan ... It has been a while, and our last few submissions weren't processed. We would like to have that done soon, if possible, as we plan a publication soon. Thank you so much

@liamdugan

Copy link
Copy Markdown
Owner

@MohamedMady19 I approved this to run, do you want me to close #169 and #167 ?

@github-actions

github-actions Bot commented Jul 6, 2026

Copy link
Copy Markdown

Eval run succeeded! Link to run: link

Here are the results of the submission(s):

DeBERTa-ConPara-v2.14

Release date: 2026-06-28

I've committed detailed results of this detector's performance on the test set to this PR.

On the RAID dataset as a whole (aggregated across all generation models, domains, decoding strategies, repetition penalties, and adversarial attacks), it achieved an AUROC of 90.96 and a TPR of 85.75% at FPR=5% and 71.64% at FPR=1%.
Without adversarial attacks, it achieved AUROC of 97.46 and a TPR of 98.67% at FPR=5% and 87.46% at FPR=1%.

DeBERTa-ConPara-v2.13

Release date: 2026-06-28

I've committed detailed results of this detector's performance on the test set to this PR.

On the RAID dataset as a whole (aggregated across all generation models, domains, decoding strategies, repetition penalties, and adversarial attacks), it achieved an AUROC of 97.15 and a TPR of 97.78% at FPR=5% and 93.86% at FPR=1%.
Without adversarial attacks, it achieved AUROC of 96.86 and a TPR of 97.40% at FPR=5% and 93.23% at FPR=1%.

DeBERTa-ConPara-v2.12

Release date: 2026-06-28

I've committed detailed results of this detector's performance on the test set to this PR.

On the RAID dataset as a whole (aggregated across all generation models, domains, decoding strategies, repetition penalties, and adversarial attacks), it achieved an AUROC of 92.49 and a TPR of 94.87% at FPR=5% and 75.10% at FPR=1%.
Without adversarial attacks, it achieved AUROC of 92.62 and a TPR of 95.04% at FPR=5% and 76.05% at FPR=1%.

If all looks well, a maintainer will come by soon to merge this PR and your entry/entries will appear on the leaderboard. If you need to make any changes, feel free to push new commits to this PR. Thanks for submitting to RAID!

@MohamedMady19

Copy link
Copy Markdown
Author

@MohamedMady19 I approved this to run, do you want me to close #169 and #167 ?

@liamdugan yes please close #167 and #169, they're superseded by #172. Thank you for merging!

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.

2 participants