AVLLMs encode audio semantics in middle layers but suppress them in final text outputs when audio conflicts with vision, due to training that largely inherits from vision-language base models.
Mmevalpro: Calibrating multimodal benchmarks towards trustworthy and efficient evaluation.ArXiv, abs/2407.00468
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
dataset 1
other 1
citation-polarity summary
fields
cs.AI 2years
2026 2representative citing papers
Hallucinations in multimodal reasoners cluster at high-entropy pivot tokens due to intermediate-layer visual anchoring failure, and V-STAR mitigates this with attention rewards plus forced reflection.
citing papers explorer
-
Do Audio-Visual Large Language Models Really See and Hear?
AVLLMs encode audio semantics in middle layers but suppress them in final text outputs when audio conflicts with vision, due to training that largely inherits from vision-language base models.
-
Cognitive Pivot Points and Visual Anchoring: Unveiling and Rectifying Hallucinations in Multimodal Reasoning Models
Hallucinations in multimodal reasoners cluster at high-entropy pivot tokens due to intermediate-layer visual anchoring failure, and V-STAR mitigates this with attention rewards plus forced reflection.