Analyzing the Mechanism of Attention Collapse in VGGT from a Dynamics Perspective

• URL: http://arxiv.org/abs/2512.21691v1
• Date: Thu, 25 Dec 2025 14:34:27 GMT
• Title: Analyzing the Mechanism of Attention Collapse in VGGT from a Dynamics Perspective
• Authors: Huan Li, Longjun Luo, Yuling Shi, Xiaodong Gu,
• Abstract summary: Visual Geometry Grounded Transformer (VGGT) delivers state-of-the-art feed-forward 3D reconstruction.<n>Its global self-attention layer suffers from a drastic collapse phenomenon when the input sequence exceeds a few hundred frames.<n>We establish a rigorous mathematical explanation of the collapse by viewing the global-attention as a degenerate diffusion process.
• Score: 13.434698786044107
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Visual Geometry Grounded Transformer (VGGT) delivers state-of-the-art feed-forward 3D reconstruction, yet its global self-attention layer suffers from a drastic collapse phenomenon when the input sequence exceeds a few hundred frames: attention matrices rapidly become near rank-one, token geometry degenerates to an almost one-dimensional subspace, and reconstruction error accumulates super-linearly.In this report,we establish a rigorous mathematical explanation of the collapse by viewing the global-attention iteration as a degenerate diffusion process.We prove that,in VGGT, the token-feature flow converges toward a Dirac-type measure at a $O(1/L)$ rate, where $L$ is the layer index, yielding a closed-form mean-field partial differential equation that precisely predicts the empirically observed rank profile.The theory quantitatively matches the attention-heat-map evolution and a series of experiments outcomes reported in relevant works and explains why its token-merging remedy -- which periodically removes redundant tokens -- slows the effective diffusion coefficient and thereby delays collapse without additional training.We believe the analysis provides a principled lens for interpreting future scalable 3D-vision transformers,and we highlight its potential for multi-modal generalization.

Related papers

• Plug-and-Play Diffusion Meets ADMM: Dual-Variable Coupling for Robust Medical Image Reconstruction [45.25461515976432]
  Plug-and-Play diffusion prior (DP) frameworks have emerged as a powerful paradigm for imaging reconstruction.<n>We present a novel approach to resolving bias-hallucination trade-off, achieving state-of-the-art gradients with significantly accelerated convergence.
  arXiv  Detail & Related papers  (2026-02-26T16:58:43Z)
• From Circuits to Dynamics: Understanding and Stabilizing Failure in 3D Diffusion Transformers [25.11520870904882]
  3D diffusion transformers exhibit a catastrophic mode of failure.<n>We call this phenomenon Meltdown.<n>We introduce PowerRemap, a test-time control that stabilizes sparse point-cloud conditioning.
  arXiv  Detail & Related papers  (2026-02-11T18:42:05Z)
• TensorLens: End-to-End Transformer Analysis via High-Order Attention Tensors [53.891337639229285]
  We introduce attentionLens, a novel formulation that captures the entire transformer as a single, input-dependent linear operator expressed through a high-order attention-interaction connection.<n>Our experiments demonstrate that the attention tensor can serve as a powerful foundation for developing tools aimed at interpretability and model understanding.
  arXiv  Detail & Related papers  (2026-01-25T19:21:25Z)
• Latent Object Permanence: Topological Phase Transitions, Free-Energy Principles, and Renormalization Group Flows in Deep Transformer Manifolds [0.5729426778193398]
  We study the emergence of multi-step reasoning in deep Transformer language models through a geometric and statistical-physics lens.<n>We formalize the forward pass as a discrete coarse-graining map and relate the appearance of stable "concept basins" to fixed points of this renormalization-like dynamics.<n>The resulting low-entropy regime is characterized by a spectral tail collapse and by the formation of transient, reusable object-like structures in representation space.
  arXiv  Detail & Related papers  (2026-01-16T23:11:02Z)
• SIGMA: Scalable Spectral Insights for LLM Collapse [51.863164847253366]
  We introduce SIGMA (Spectral Inequalities for Gram Matrix Analysis), a unified framework for model collapse.<n>By utilizing benchmarks that deriving and deterministic bounds on the matrix's spectrum, SIGMA provides a mathematically grounded metric to track the contraction of the representation space.<n>We demonstrate that SIGMA effectively captures the transition towards states, offering both theoretical insights into the mechanics of collapse.
  arXiv  Detail & Related papers  (2026-01-06T19:47:11Z)
• Random-Matrix-Induced Simplicity Bias in Over-parameterized Variational Quantum Circuits [72.0643009153473]
  We show that expressive variational ansatze enter a Haar-like universality class in which both observable expectation values and parameter gradients concentrate exponentially with system size.<n>As a consequence, the hypothesis class induced by such circuits collapses with high probability to a narrow family of near-constant functions.<n>We further show that this collapse is not unavoidable: tensor-structured VQCs, including tensor-network-based and tensor-hypernetwork parameterizations, lie outside the Haar-like universality class.
  arXiv  Detail & Related papers  (2026-01-05T08:04:33Z)
• Hierarchical Attention for Sparse Volumetric Anomaly Detection in Subclinical Keratoconus [0.0]
  hierarchical architectures achieve 21-23% higher sensitivity and specificity, particularly in the difficult subclinical regime.<n>Mechanistic analyses indicate that this advantage arises from spatial scale alignment.<n>Subclinical cases require longer spatial integration than healthy or overtly pathological volumes.
  arXiv  Detail & Related papers  (2025-12-03T01:20:13Z)
• Learning Time-Varying Graphs from Incomplete Graph Signals [1.7430416823420511]
  We develop an efficient Alternating Direction Multiplier algorithm for solving the problem of imputing missing data from a graph.<n>We prove that the proposed ADMM scheme converges to and we derive a stationary point.
  arXiv  Detail & Related papers  (2025-10-19T11:12:13Z)
• Tracing the Representation Geometry of Language Models from Pretraining to Post-training [22.18942718274405]
  We take a spectral approach to investigate the geometry of learned representations across pretraining and post-training.<n>We uncover a consistent non-monotonic sequence of three geometric phases during autoregressive pretraining.<n>Post-training further transforms geometry: SFT and DPO drive "entropy-seeking" dynamics to integrate specific instructional or preferential data.
  arXiv  Detail & Related papers  (2025-09-27T00:46:29Z)
• Beyond Progress Measures: Theoretical Insights into the Mechanism of Grokking [50.465604300990904]
  Grokking refers to the abrupt improvement in test accuracy after extended overfitting.<n>In this work, we investigate the grokking mechanism underlying the Transformer in the task of prime number operations.
  arXiv  Detail & Related papers  (2025-04-04T04:42:38Z)
• Mind the Gap: a Spectral Analysis of Rank Collapse and Signal Propagation in Attention Layers [3.686808512438363]
  Alternatives to softmax-based attention are being due to its tendency to hinder effective information flow.<n>We conduct a rigorous analysis to uncover a spectral gap between the two largest singular gradients of the attention matrix.<n>We propose a novel simple practical solution to rank collapse in width by removing the outlier(s)
  arXiv  Detail & Related papers  (2024-10-10T10:34:18Z)
• Rank Collapse Causes Over-Smoothing and Over-Correlation in Graph Neural Networks [3.566568169425391]
  We show that with increased depth, node representations become dominated by a low-dimensional subspace that depends on the aggregation function but not on the feature transformations. For all aggregation functions, the rank of the node representations collapses, resulting in over-smoothing for particular aggregation functions.
  arXiv  Detail & Related papers  (2023-08-31T15:22:31Z)
• Convergence of mean-field Langevin dynamics: Time and space discretization, stochastic gradient, and variance reduction [49.66486092259376]
  The mean-field Langevin dynamics (MFLD) is a nonlinear generalization of the Langevin dynamics that incorporates a distribution-dependent drift. Recent works have shown that MFLD globally minimizes an entropy-regularized convex functional in the space of measures. We provide a framework to prove a uniform-in-time propagation of chaos for MFLD that takes into account the errors due to finite-particle approximation, time-discretization, and gradient approximation.
  arXiv  Detail & Related papers  (2023-06-12T16:28:11Z)
• Implicit Bias of Gradient Descent for Logistic Regression at the Edge of Stability [69.01076284478151]
  In machine learning optimization, gradient descent (GD) often operates at the edge of stability (EoS) This paper studies the convergence and implicit bias of constant-stepsize GD for logistic regression on linearly separable data in the EoS regime.
  arXiv  Detail & Related papers  (2023-05-19T16:24:47Z)
• Revisiting Over-smoothing in BERT from the Perspective of Graph [111.24636158179908]
  Recently over-smoothing phenomenon of Transformer-based models is observed in both vision and language fields. We find that layer normalization plays a key role in the over-smoothing issue of Transformer-based models. We consider hierarchical fusion strategies, which combine the representations from different layers adaptively to make the output more diverse.
  arXiv  Detail & Related papers  (2022-02-17T12:20:52Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.