AVLLMs store integrated audio-visual information mainly in a distinct subset of sink tokens called cross-modal sink tokens, which can be leveraged for training-free hallucination mitigation.
Some modalities are more equal than others: Decoding and architecting multi- modal integration in mllms.arXiv preprint arXiv:2511.22826
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3verdicts
UNVERDICTED 3representative citing papers
Coordinated multi-modal typographic attacks on MLLMs achieve 83.43% success rate versus 34.93% for single-modality attacks.
IPPg embeds text into images to reduce multimodal model inference costs by 35.8-91% with competitive accuracy on many VQA and code benchmarks.
citing papers explorer
-
Probing Cross-modal Information Hubs in Audio-Visual LLMs
AVLLMs store integrated audio-visual information mainly in a distinct subset of sink tokens called cross-modal sink tokens, which can be leveraged for training-free hallucination mitigation.
-
A Systematic Study of Cross-Modal Typographic Attacks on Audio-Visual Reasoning
Coordinated multi-modal typographic attacks on MLLMs achieve 83.43% success rate versus 34.93% for single-modality attacks.
-
Token-Efficient Multimodal Reasoning via Image Prompt Packaging
IPPg embeds text into images to reduce multimodal model inference costs by 35.8-91% with competitive accuracy on many VQA and code benchmarks.