arxiv:2609.03586

The Attention Triangle in Audio-Video Models

Published on Sep 3

· Submitted by

Sagi Polaczek on Sep 7

Tel Aviv University

Upvote

Authors:

Abstract

Audio-video diffusion models exhibit bidirectional semantic leakage through cross-modal attention pathways, which can be diagnosed via attention-derived signals and mitigated through inference-time alignment interventions.

Generated by thinkingmachines/Inkling-Small

Audio-video diffusion models rely on cross-modal attention to coordinate text, sound, and visual content, yet this same mechanism can introduce subtle and systematic semantic leakage. We study these models by probing and analyzing the ``attention triangle,'' comprising the three cross-attention edges connecting the text, audio, and video streams, and examine how semantic information is routed across modalities during generation. Our analysis reveals that routing along the audio-video edge is bidirectional: audio can influence video generation, while video can influence audio generation. This edge is shaped by biases encoded in the model's parameters and emerges as a major contributor to leakage: when prompts are in tension with learned priors, cross-modal interactions may override the intended conditioning and reroute semantics toward visually canonical but incorrect outcomes. These effects suggest that semantic artifacts arise not merely from attention spreading beyond its intended target, but from structured, bias-driven interactions along specific pathways. Building on this perspective, we extract attention-derived signals that expose how semantics are distributed and grounded across modalities, and use them as a diagnostic tool to both analyze and deliberately incur leakage under controlled conditions. This enables us to probe the internal dynamics of cross-modal routing and isolate the role of individual interactions. We further leverage these signals to guide inference-time interventions that encourage more consistent cross-modal alignment. Extensive experiments support our analysis and demonstrate improved semantic grounding while preserving generation quality.

View arXiv page View PDF Project page GitHub 2 Add to collection

Community

SagiPolaczek

Paper submitter about 20 hours ago

Audio-video diffusion models rely on cross-modal attention to coordinate text, sound, and visual content, yet this same mechanism can introduce subtle and systematic semantic leakage.
We study these models by probing and analyzing the ``attention triangle,'' comprising the three cross-attention edges connecting the text, audio, and video streams, and examine how semantic information is routed across modalities during generation.
Our analysis reveals that routing along the audio-video edge is bidirectional: audio can influence video generation, while video can influence audio generation.
This edge is shaped by biases encoded in the model's parameters and emerges as a major contributor to leakage: when prompts are in tension with learned priors, cross-modal interactions may override the intended conditioning and reroute semantics toward visually canonical but incorrect outcomes.
These effects suggest that semantic artifacts arise not merely from attention spreading beyond its intended target, but from structured, bias-driven interactions along specific pathways.
Building on this perspective, we extract attention-derived signals that expose how semantics are distributed and grounded across modalities, and use them as a diagnostic tool to both analyze and deliberately incur leakage under controlled conditions.
This enables us to probe the internal dynamics of cross-modal routing and isolate the role of individual interactions. We further leverage these signals to guide inference-time interventions that encourage more consistent cross-modal alignment.
Extensive experiments support our analysis and demonstrate improved semantic grounding while preserving generation quality.

🌐 visit our project page at: https://sagipolaczek.github.io/The-Attention-Triangle/