MM-DeceptionBench — leaderboard

Multimodal deception benchmark for vision-language models with image-grounded cases covering sycophancy, sandbagging, bluffing, obfuscation, deliberate omission, and fabrication.

Source: huggingface.co.

Interactive version: theaggregate.ai/benchmark?slug=mm-deceptionbench · How the rankings work · Data refreshed daily, snapshot 2026-07-22.