Add flaird-gated-multitask

[yusrpro9]

No description provided.

[github-actions]

Eval run succeeded! Link to run: link

Here are the results of the submission(s):

flaird-gated-multitask

Release date: 2026-08-24

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 99.80 and a TPR of 99.04% at FPR=5% and 97.53% at FPR=1%.
Without adversarial attacks, it achieved AUROC of 99.98 and a TPR of 99.92% at FPR=5% and 99.77% 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!

[liamdugan]

@yusrpro9 congrats on the strong performance. Let me know if you'd like me to merge this into the leaderboard.

[yusrpro9]

Dear Liam, Thank you for reaching out. I have trained a few additional models and am currently waiting for their evaluation results. I plan to evaluate them this week, so I would prefer to wait for the final results before deciding which model to submit to the leaderboard. Best regards, Yusr

…

On Fri, 2 Oct 2026 at 9:58 PM Liam Dugan ***@***.***> wrote: *liamdugan* left a comment (liamdugan/raid#192) <#192 (comment)> @yusrpro9 <https://github.com/yusrpro9> congrats on the strong performance. Let me know if you'd like me to merge this into the leaderboard. — Reply to this email directly, view it on GitHub <#192?email_source=notifications&email_token=CEHSTESFW7MQ4GVY5CIRPQL5R73HDA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTKOJVHEZTIMRTGA3KM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLDGN5XXIZLSL5RWY2LDNM#issuecomment-5959342306>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/CEHSTESK5AWGENBXPUTVXZT5R73HDAVCNFSNUABFKJSXA33TNF2G64TZHM4DAMBQHAZTANJQHNEXG43VMU5TKMRTGUZTKNRSGM3KC5QC> . Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS <https://github.com/notifications/mobile/ios/CEHSTEUDNETVGKYUPHAD7X35R73HDA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTKOJVHEZTIMRTGA3KM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJKTGN5XXIZLSL5UW64Y> and Android <https://github.com/notifications/mobile/android/CEHSTEQ4H7HD7SIM6EVLA2D5R73HDA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTKOJVHEZTIMRTGA3KM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLTGN5XXIZLSL5QW4ZDSN5UWI>. Download it today! You are receiving this because you were mentioned.Message ID: ***@***.***>

[liamdugan]

No worries! Just ping me when you have those submissions ready

[yusrpro9]

Dear Liam, I have uploaded several models. I would appreciate it if you could evaluate them as soon as possible and upload the results to the leaderboard, as I need the evaluation results to proceed with my research. Thank you. .

…

On Sat, 3 Oct 2026 at 12:23 AM Liam Dugan ***@***.***> wrote: *liamdugan* left a comment (liamdugan/raid#192) <#192 (comment)> No worries! Just ping me when you have those submissions ready — Reply to this email directly, view it on GitHub <#192?email_source=notifications&email_token=CEHSTETSMKHYZFD7KRV22LD5SAMF7A5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTKOJWGE3DSNRQGQ2KM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLDGN5XXIZLSL5RWY2LDNM#issuecomment-5961696044>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/CEHSTETYFX6QNHG2CSOS5ZT5SAMF7AVCNFSNUABFKJSXA33TNF2G64TZHM4DAMBQHAZTANJQHNEXG43VMU5TKMRTGUZTKNRSGM3KC5QC> . Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS <https://github.com/notifications/mobile/ios/CEHSTEXJHCEIEEUQ6C6EGGL5SAMF7A5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTKOJWGE3DSNRQGQ2KM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJKTGN5XXIZLSL5UW64Y> and Android <https://github.com/notifications/mobile/android/CEHSTEWONLQCOIXJGYFO5AL5SAMF7A5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTKOJWGE3DSNRQGQ2KM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLTGN5XXIZLSL5QW4ZDSN5UWI>. Download it today! You are receiving this because you were mentioned.Message ID: ***@***.***>

[github-actions]

It looks like this eval run failed. Please check the workflow logs to see what went wrong, then push a new commit to your PR to rerun the eval.

[liamdugan]

Hi @yusrpro9 it seems like there was an error with your submission. Are you sure your JSON is formatted properly?

Also, please refrain from submitting many different configurations / grid searches to the RAID leaderboard. From what I can tell you have a 2x3 grid search here with concatenation vs. attention and multitask vs. single-task and frozen vs. non-frozen. Please grid search on your own development set and then just submit your final model to the leaderboard.

[liamdugan]

@yusrpro9 I see you re-requested a deployment. However, I won't run eval on this submission until you remove the other comparisons. Please submit only one model.

[yusrpro9]

The JSON files formatting error has been resolved. The models have different architectures and training mechanisms, and I need to evaluate them all. You can upload only the best model you obtain to the leaderboard under the name flaird.

…

On Tue, 6 Oct 2026 at 5:06 PM Liam Dugan ***@***.***> wrote: *liamdugan* left a comment (liamdugan/raid#192) <#192 (comment)> Hi @yusrpro9 <https://github.com/yusrpro9> it seems like there was an error with your submission. Are you sure your JSON is formatted properly? Also, please refrain from submitting many different configurations / grid searches to the RAID leaderboard. From what I can tell you have a 2x3 grid search here with concatenation vs. attention and multitask vs. single-task and frozen vs. non-frozen. Please grid search on your own development set and then just submit your final model to the leaderboard. — Reply to this email directly, view it on GitHub <#192?email_source=notifications&email_token=CEHSTEU4GYIV43S45NZXZKT5ST37RA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTMMBRHAYDAMRVGE22M4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLDGN5XXIZLSL5RWY2LDNM#issuecomment-6018002515>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/CEHSTESNK4LTIKJKMEIPFL35ST37RAVCNFSNUABFKJSXA33TNF2G64TZHM4DAMBQHAZTANJQHNEXG43VMU5TKMRTGUZTKNRSGM3KC5QC> . Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS <https://github.com/notifications/mobile/ios/CEHSTEQ663JZWKR5T5IQYCT5ST37RA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTMMBRHAYDAMRVGE22M4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJKTGN5XXIZLSL5UW64Y> and Android <https://github.com/notifications/mobile/android/CEHSTERMTXQHK23F77GJOJD5ST37RA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTMMBRHAYDAMRVGE22M4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLTGN5XXIZLSL5QW4ZDSN5UWI>. Download it today! You are receiving this because you were mentioned.Message ID: ***@***.***>

[liamdugan]

@yusrpro9 frozen vs. non frozen & multi-task vs. single task are not different architectures. Please evaluate these models on your dev set and submit only one to the leaderboard.