Video MLLMs show an audio-visual Clever Hans effect relying on visual-acoustic correlations rather than audio verification; Thud interventions diagnose it and a 10K-sample preference alignment improves intervention performance by 28 points.
Qwen3.5-omni technical report, 2026
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
Omni-Encoder unifies visual and audio encoding at symmetrical 25 fps using a Transformer with three new components, yielding gains on fine-grained motion tasks while matching baselines on audio-visual benchmarks.
citing papers explorer
-
When Vision Speaks for Sound
Video MLLMs show an audio-visual Clever Hans effect relying on visual-acoustic correlations rather than audio verification; Thud interventions diagnose it and a 10K-sample preference alignment improves intervention performance by 28 points.
-
OmniEncoder: See, Hear, and Feel Continuous Motion Like Humans With One Encoder
Omni-Encoder unifies visual and audio encoding at symmetrical 25 fps using a Transformer with three new components, yielding gains on fine-grained motion tasks while matching baselines on audio-visual benchmarks.