Tag
This paper studies how audio and visual information flow inside Audio-Visual Large Language Models (AVLLMs), revealing that AVLLMs follow sequential or parallel routing depending on input configuration, and that some tokens can be discarded after information transfer for efficiency.
This paper proposes a framework to evaluate and improve faithfulness of chain-of-thought reasoning by controlling information flow, using entropy-based, KL-divergence, and gradient-based diagnostics, and introduces training interventions (attention masking, gradient masking, adversarial perturbations) that make reasoning more transparent and reduce shortcut reliance.
Recommend two free and open-source tools: TrendRadar (Chinese region, 58k+ stars) and Horizon (English region, 4.6k+ stars). They can automatically capture hot topics across the entire internet, use AI to filter high-value content, generate translated and analytical briefs, and push them with one click to multiple platforms.
This paper proposes Diffusion-Adaptive Routing (DAR), a learnable, timestep-adaptive residual replacement that improves cross-layer information flow in Diffusion Transformers, leading to significant training acceleration and quality improvements.
This paper investigates how semantic information is distributed across textual tokens in text-to-image models, finding that information concentration and cross-item interactions significantly affect image generation alignment. The authors use patching techniques to demonstrate that simple encoding-stage interventions can improve alignment quality.