Image de couverture de livre

A Distributional View for Visual Mechanistic Interpretability: KL-Minimal Soft-Constraint Principle

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Zhou, Guancheng, Luo, Yisi, He, Zhengfu, Jin, Zhenyu, Ge, Xuyang, Shu, Wentao, Meng, Deyu, Qiu, Xipeng
Format:	Preprint
Publié:	2026
Sujets:	Computer Vision and Pattern Recognition Artificial Intelligence
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Dimensional Collapse in Transformer Attention Outputs: A Challenge for Sparse Dictionary Learning
par: Wang, Junxuan, et autres
Publié: (2025)

Towards Universality: Studying Mechanistic Similarity Across Language Model Architectures
par: Wang, Junxuan, et autres
Publié: (2024)

Dictionary Learning Improves Patch-Free Circuit Discovery in Mechanistic Interpretability: A Case Study on Othello-GPT
par: He, Zhengfu, et autres
Publié: (2024)

Tracing the Thought of a Grandmaster-level Chess-Playing Transformer
par: Lin, Rui, et autres
Publié: (2026)

Automatically Identifying Local and Global Circuits with Linear Computation Graphs
par: Ge, Xuyang, et autres
Publié: (2024)

Evolution of Concepts in Language Model Pre-Training
par: Ge, Xuyang, et autres
Publié: (2025)

Towards Understanding the Nature of Attention with Low-Rank Sparse Decomposition
par: He, Zhengfu, et autres
Publié: (2025)

Compressive Imaging Reconstruction via Tensor Decomposed Multi-Resolution Grid Encoding
par: Jin, Zhenyu, et autres
Publié: (2025)

Revisiting Nonlocal Self-Similarity from Continuous Representation
par: Luo, Yisi, et autres
Publié: (2024)

Continuous Representation Methods, Theories, and Applications: An Overview and Perspectives
par: Luo, Yisi, et autres
Publié: (2025)

Llama Scope: Extracting Millions of Features from Llama-3.1-8B with Sparse Autoencoders
par: He, Zhengfu, et autres
Publié: (2024)

Unveiling the Mechanism of Continuous Representation Full-Waveform Inversion: A Wave Based Neural Tangent Kernel Framework
par: Chen, Ruihua, et autres
Publié: (2026)

NeurTV: Total Variation on the Neural Domain
par: Luo, Yisi, et autres
Publié: (2024)

Cross-Frequency Implicit Neural Representation with Self-Evolving Parameters
par: Yu, Chang, et autres
Publié: (2025)

Beyond Low-rankness: Guaranteed Matrix Recovery via Modified Nuclear Norm
par: Peng, Jiangjun, et autres
Publié: (2025)

Deciphering Neural Reparameterized Full-Waveform Inversion with Neural Sensitivity Kernel and Wave Tangent Kernel
par: Chen, Ruihua, et autres
Publié: (2026)

Spatial Information Bottleneck for Interpretable Visual Recognition
par: Shu, Kaixiang, et autres
Publié: (2025)

Neural Approximation and Its Applications
par: Wu, Wei-Hao, et autres
Publié: (2026)

Neural Uncertainty Principle: A Unified View of Adversarial Fragility and LLM Hallucination
par: Zhang, Dong-Xiao, et autres
Publié: (2026)

Generate Point Clouds with Multiscale Details from Graph-Represented Structures
par: Yang, Ximing, et autres
Publié: (2021)

Simultaneous Swap Regret Minimization via KL-Calibration
par: Luo, Haipeng, et autres
Publié: (2025)

Faithful and Stable Neuron Explanations for Trustworthy Mechanistic Interpretability
par: Yan, Ge, et autres
Publié: (2025)

Principled Out-of-Distribution Generalization via Simplicity
par: Ge, Jiawei, et autres
Publié: (2025)

Can AI Assistants Know What They Don't Know?
par: Cheng, Qinyuan, et autres
Publié: (2024)

Do Agents Think Deeper? A Mechanistic Investigation of Layer-Wise Dynamics in Sequential Planning
par: Cui, Zhenyu, et autres
Publié: (2026)

Making Large Language Models Better Reasoners with Orchestrated Streaming Experiences
par: Liu, Xiangyang, et autres
Publié: (2025)

Soft-Evidence Fused Graph Neural Network for Cancer Driver Gene Identification across Multi-View Biological Graphs
par: Chen, Bang, et autres
Publié: (2025)

Interpretability in Parameter Space: Minimizing Mechanistic Description Length with Attribution-based Parameter Decomposition
par: Braun, Dan, et autres
Publié: (2025)

TRG-Net: An Interpretable and Controllable Rain Generator
par: Pang, Zhiqiang, et autres
Publié: (2024)

A Unified Dual Consensus Approach to Distributed Optimization with Globally-Coupled Constraints
par: Liu, Zixuan, et autres
Publié: (2025)

Chain-of-Restoration: Multi-Task Image Restoration Models are Zero-Shot Step-by-Step Universal Image Restorers
par: Cao, Jin, et autres
Publié: (2024)

GAOKAO-MM: A Chinese Human-Level Benchmark for Multimodal Models Evaluation
par: Zong, Yi, et autres
Publié: (2024)

Differentially Private Distribution Release of Gaussian Mixture Models via KL-Divergence Minimization
par: Liu, Hang, et autres
Publié: (2025)

A Novel Control Method of Sugar Boiling Based on Model‐Free Adaptive Control and Neural Networks
par: Guancheng Lu, et autres
Publié: (2025)

Understanding Multimodal LLMs: the Mechanistic Interpretability of Llava in Visual Question Answering
par: Yu, Zeping, et autres
Publié: (2024)

Limiting Behavior of Constraint Minimizers for Inhomogeneous Fractional Schrödinger Equations
par: Zhang, Hongfei, et autres
Publié: (2023)

A Novel Robot Hand with Hoeckens Linkages and Soft Phalanges for Scooping and Self-Adaptive Grasping in Environmental Constraints
par: Guo, Wentao, et autres
Publié: (2025)

MINAR: Mechanistic Interpretability for Neural Algorithmic Reasoning
par: He, Jesse, et autres
Publié: (2026)

Improving Memory Efficiency for Training KANs via Meta Learning
par: Zhao, Zhangchi, et autres
Publié: (2025)

Understanding the Generalization of Bilevel Programming in Hyperparameter Optimization: A Tale of Bias-Variance Decomposition
par: Zhou, Yubo, et autres
Publié: (2026)