Trying luminar_classifier_RAID_none_PrismAI

[TheItCrOw]

After the evaluation changes, testing new luminar model instances. I'll probably add more.

[github-actions[bot]]

Eval run succeeded! Link to run: link

Here are the results of the submission(s):

e5-small-lora

Release date: 2024-11-07

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 96.85 and a TPR of 85.69% at FPR=5% and 73.08% at FPR=1%. Without adversarial attacks, it achieved AUROC of 98.61 and a TPR of 93.87% at FPR=5% and 82.81% at FPR=1%.

luminar_classifier_RAID_none_PrismAI

Release date: 2025-05-17

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

[!WARNING] Failed to find threshold values that achieve False Positive Rate(s): (['5%', '1%']) on all domains. This submission will not appear in the main leaderboard for those FPR values; it will only be visible within the splits in which the target FPR was achieved.

LLMDet

Release date: 2023-05-24

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 62.90 and a TPR of 26.70% at FPR=5% and 14.91% at FPR=1%. Without adversarial attacks, it achieved AUROC of 65.92 and a TPR of 33.40% at FPR=5% and 20.16% at FPR=1%.

Luminar

Release date: 2025-05-17

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

[!WARNING] Failed to find threshold values that achieve False Positive Rate(s): (['5%', '1%']) on all domains. This submission will not appear in the main leaderboard for those FPR values; it will only be visible within the splits in which the target FPR was achieved.

SpeedAI

Release date: 2025-05-08

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 99.85 and a TPR of 99.62% at FPR=5% and 98.55% at FPR=1%. Without adversarial attacks, it achieved AUROC of 99.91 and a TPR of 99.86% at FPR=5% and 99.55% at FPR=1%.

It's AI

Release date: 2025-04-01

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 94.27 and a TPR of 94.15% at FPR=5% and 89.36% at FPR=1%. Without adversarial attacks, it achieved AUROC of 95.21 and a TPR of 95.75% at FPR=5% and 92.22% at FPR=1%.

RoBERTa-base (GPT2)

Release date: 2019-08-24

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 72.29 and a TPR of 51.77% at FPR=5% and 34.57% at FPR=1%. Without adversarial attacks, it achieved AUROC of 74.95 and a TPR of 59.19% at FPR=5% and 39.83% at FPR=1%.

RoBERTa (ChatGPT)

Release date: 2023-01-18

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 60.26 and a TPR of 26.64% at FPR=5% and 19.63% at FPR=1%. Without adversarial attacks, it achieved AUROC of 77.29 and a TPR of 42.52% at FPR=5% and 32.84% at FPR=1%.

Desklib

Release date: 2024-10-03

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 94.91 and a TPR of 83.76% at FPR=5% and 68.22% at FPR=1%. Without adversarial attacks, it achieved AUROC of 97.39 and a TPR of 92.38% at FPR=5% and 86.27% at FPR=1%.

RoBERTa-large (GPT2)

Release date: 2019-08-24

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 69.28 and a TPR of 50.70% at FPR=5% and 34.67% at FPR=1%. Without adversarial attacks, it achieved AUROC of 71.60 and a TPR of 55.73% at FPR=5% and 38.83% at FPR=1%.

SuperAnnotate AI Detector

Release date: 2024-10-27

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 88.87 and a TPR of 64.87% at FPR=5% and 38.87% at FPR=1%. Without adversarial attacks, it achieved AUROC of 90.96 and a TPR of 70.34% at FPR=5% and 44.53% at FPR=1%.

GLTR

Release date: 2019-06-10

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 70.90 and a TPR of 51.48% at FPR=5% and 36.48% at FPR=1%. Without adversarial attacks, it achieved AUROC of 72.68 and a TPR of 59.69% at FPR=5% and 45.57% at FPR=1%.

Desklib AI Text Detector v1.01

Release date: 2025-02-16

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 94.83 and a TPR of 91.17% at FPR=5% and 76.47% at FPR=1%. Without adversarial attacks, it achieved AUROC of 97.34 and a TPR of 94.87% at FPR=5% and 89.11% at FPR=1%.

Binoculars

Release date: 2024-01-22

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

[!WARNING] No aggregate score across all settings is reported here as some domains/generator models/decoding strategies/repetition penalties/adversarial attacks were not included in the submission. This submission will not appear in the main leaderboard; it will only be visible within the splits in which all samples were evaluated. Without adversarial attacks, it achieved AUROC of 84.40 and a TPR of 78.98% at FPR=5% and 69.54% at FPR=1%.

RADAR

Release date: 2023-07-07

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 81.92 and a TPR of 63.91% at FPR=5% and 43.12% at FPR=1%. Without adversarial attacks, it achieved AUROC of 81.89 and a TPR of 65.61% at FPR=5% and 48.09% at FPR=1%.

Gaussian Extreme

Release date: 2025-05-17

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

[!WARNING] Failed to find threshold values that achieve False Positive Rate(s): (['5%', '1%']) on all domains. This submission will not appear in the main leaderboard for those FPR values; it will only be visible within the splits in which the target FPR was achieved.

FastDetectGPT

Release date: 2023-10-08

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

[!WARNING] No aggregate score across all settings is reported here as some domains/generator models/decoding strategies/repetition penalties/adversarial attacks were not included in the submission. This submission will not appear in the main leaderboard; it will only be visible within the splits in which all samples were evaluated.

[!WARNING] No aggregate score across all non-adversarial settings is reported here as some domains/generator models/decoding strategies/repetition penalties were not included in the submission.

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!