Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Deception Detection research benchmarks – Page 2 · SOTA2 Research

Deception Detection

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

Real-life Trial Deception (Honest Leave-One-Speaker-Out)

Self-hosted features → gradient-boosting

67.8Accuracy

6

Jun 23, 2026

93.5Response AUROC

6

May 20, 2026

1AUROC

4

May 28, 2026

DeceptArena (test)

0.927False Assertion Score

4

Mar 17, 2026

7 non-control benchmark families Mean across targets

91.6AUROC

3

Jun 17, 2026

7 non-control benchmark families GEMMA4-26B-A4B-IT

91.1AUROC

3

Jun 17, 2026

7 non-control benchmark families GPT-OSS-20B

0.92AUROC

3

Jun 17, 2026

Natural Instructed Deception (held-out facts)

100Orientation Accuracy

3

Jun 17, 2026

Synthetic Deception facts (held-out)

100Orientation Accuracy

3

Jun 17, 2026

56.41Accuracy

3

Jun 11, 2026

Liars' Bench Harm-Pressure Knowledge (test)

5-Layer Ensemble

0.91AUROC

3

Apr 16, 2026

Liars' Bench Insider Trading (test)

5-Layer Ensemble

0.953AUROC

3

Apr 16, 2026

Liars' Bench Convincing Game (test)

1AUROC

3

Apr 16, 2026

Liars' Bench Harm-Pressure Choice (test)

0.949AUROC

3

Apr 16, 2026

Liars' Bench Instructed Deception (test)

0.939AUROC

3

Apr 16, 2026

DOLOs, Bag-of-Lies, and MU3D Average

60.11Accuracy

3

Mar 26, 2026

57.24Accuracy

3

Mar 26, 2026

Bag-of-Lies (test)

59.62Accuracy

3

Mar 26, 2026

63.46Accuracy

3

Mar 26, 2026

Deception Detection (test)

10Behavioral Value (Acc Diff %)

3

Feb 26, 2026

CharXiv modified images removed

9Deception Rate

2

Feb 26, 2026

Browsing Broken Tools tasks

0.11Deception Rate

2

Feb 26, 2026

Agentic Coding tasks

0.17Deception Rate

2

Feb 26, 2026

Page 2 of 2

Popular datasets

Liars' Bench DOLOs Bag-of-Lies Non-control benchmark families MDPE

Follow for updates

Join SOTA2 Contact