paper-with-me

Papers

Beyond Attention Magnitude: Leveraging Inter-layer Rank Consistency for Efficient Vision-Language-Action Models

2026-03-26 · Peiju Liu, Jinming Liu, Xipeng Qiu, Xuanjing Huang arxiv

Vision-Language-Action (VLA) models excel in robotic manipulation but suffer from significant inference latency due to processing dense visual tokens. Existing token reduction methods predominantly rely on attention magnitude as a static selection. In this work, we challenge this assumption, revealing that high-attention tokens are task-dependent and can even degrade policy performance. To address this, we introduce \textbf{TIES} (\textbf{T}au-guided \textbf{I}nter-layer \textbf{E}fficient \textbf{S}election), a dynamic framework guided by inter-layer token ranking consistency. By adaptively balancing attention magnitude with ranking consistency, TIES ensures robust token selection without requiring additional training. On the CogACT + SIMPLER benchmark, TIES improves average success rates by 6\% while reducing token usage by 78\%, and demonstrate strong generalization across diverse decoders and benchmarks.

📄 PDF Abstract BibTeX arXiv:2603.24941

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

SGFormer: Single-Layer Graph Transformers with Approximation-Free Linear Complexity

2024-09-13 · Qitian Wu, Kai Yang, Hengrui Zhang, David Wipf 외

Learning representations on large graphs is a long-standing challenge due to the inter-dependence nature. Transformers recently have shown promising performance on small graphs thanks to its global attention for capturin…

Deep AttentionRepresentation Learning

VARA-TTS: Non-Autoregressive Text-to-Speech Synthesis based on Very Deep VAE with Residual Attention

2021-02-12 · Peng Liu, Yuewen Cao, Songxiang Liu, Na Hu 외

This paper proposes VARA-TTS, a non-autoregressive (non-AR) text-to-speech (TTS) model using a very deep Variational Autoencoder (VDVAE) with Residual Attention mechanism, which refines the textual-to-acoustic alignment …

Speech Synthesistext-to-speechText to SpeechText-To-Speech Synthesis

Disentangling Direction and Magnitude in Transformer Representations: A Double Dissociation Through L2-Matched Perturbation Analysis

2026-01-19 · Mangadoddi Srikar Vardhan, Lekkala Sai Teja arxiv

Transformer hidden states encode information as high-dimensional vectors, yet whether direction (orientation in representational space) and magnitude (vector norm) serve distinct functional roles remains unclear. Studyin…

Transformers Are Born Biased: Structural Inductive Biases at Random Initialization and Their Practical Consequences

2026-02-05 · Siquan Li, Yao Tong, Haonan Wang, Tianyang Hu arxiv

Transformers underpin modern large language models (LLMs) and are commonly assumed to be behaviorally unstructured at random initialization, with all meaningful preferences emerging only through large-scale training. We …

Garbage Attention in Large Language Models: BOS Sink Heads and Sink-aware Pruning

2026-01-11 · Jaewon Sok, Jewon Yeom, Seonghyeon Park, Jeongjae Park 외 arxiv

Large Language Models (LLMs) are known to contain significant redundancy, yet a systematic explanation for why certain components, particularly in higher layers, are more redundant has remained elusive. In this work, we …

Model Compression