Copertina

Depth, Not Data: An Analysis of Hessian Spectral Bifurcation

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Deng, Shenyang, Liao, Boyao, Ouyang, Zhuoli, Pang, Tianyu, Yang, Yaoqing
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Machine Learning
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Suspicious Alignment of SGD: A Fine-Grained Step Size Condition Analysis
di: Deng, Shenyang, et al.
Pubblicazione: (2026)

RMNP: Row-Momentum Normalized Preconditioning for Scalable Matrix-Based Optimization
di: Deng, Shenyang, et al.
Pubblicazione: (2026)

From Spikes to Heavy Tails: Unveiling the Spectral Evolution of Neural Networks
di: Kothapalli, Vignesh, et al.
Pubblicazione: (2024)

HTMuon: Improving Muon via Heavy-Tailed Spectral Correction
di: Pang, Tianyu, et al.
Pubblicazione: (2026)

Balancing Learning Rates Across Layers: Exact Two-Step Dynamics and Optimal Scaling in Linear Neural Networks
di: Pang, Tianyu, et al.
Pubblicazione: (2026)

KCES: Training-Free Defense for Robust Graph Neural Networks via Kernel Complexity
di: Jia, Yaning, et al.
Pubblicazione: (2025)

Spectral Insights into Data-Oblivious Critical Layers in Large Language Models
di: Liu, Xuyuan, et al.
Pubblicazione: (2025)

Model Balancing Helps Low-data Training and Fine-tuning
di: Liu, Zihang, et al.
Pubblicazione: (2024)

Hessian Spectral Analysis at Foundation Model Scale
di: Granziol, Diego, et al.
Pubblicazione: (2026)

Learning to Discover Iterative Spectral Algorithms
di: Liu, Zihang, et al.
Pubblicazione: (2026)

Learning-Zone Energy: Online Data Selection for Efficient RL Post-Training
di: Cui, Peng, et al.
Pubblicazione: (2026)

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets
di: Hsiung, Lei, et al.
Pubblicazione: (2025)

Iterative Refinement Neural Operators are Learned Fixed-Point Solvers: A Principled Approach to Spectral Bias Mitigation
di: Liu, Xiaotian, et al.
Pubblicazione: (2026)

Understanding Generalization from Embedding Dimension and Distributional Convergence
di: Yu, Junjie, et al.
Pubblicazione: (2026)

Local Hessian Spectral Filtering for Robust Intrinsic Dimension Estimation
di: Osada, Genki
Pubblicazione: (2026)

LIFT the Veil for the Truth: Principal Weights Emerge after Rank Reduction for Reasoning-Focused Supervised Fine-Tuning
di: Liu, Zihang, et al.
Pubblicazione: (2025)

Ranking-Aware Calibration for Reliable Multimodal Reinforcement Learning
di: Cui, Peng, et al.
Pubblicazione: (2026)

Eigenspectrum Analysis of Neural Networks without Aspect Ratio Bias
di: Hu, Yuanzhe, et al.
Pubblicazione: (2025)

REFORMER: A ChatGPT-Driven Data Synthesis Framework Elevating Text-to-SQL Models
di: Liu, Shenyang, et al.
Pubblicazione: (2025)

Feature Dynamics as Implicit Data Augmentation: A Depth-Decomposed View on Deep Neural Network Generalization
di: Ruan, Tianyu, et al.
Pubblicazione: (2025)

Bayesian Influence Functions for Hessian-Free Data Attribution
di: Kreer, Philipp Alexander, et al.
Pubblicazione: (2025)

Better Training Data Attribution via Better Inverse Hessian-Vector Products
di: Wang, Andrew, et al.
Pubblicazione: (2025)

A Three-regime Model of Network Pruning
di: Zhou, Yefan, et al.
Pubblicazione: (2023)

Nonparametric Data Attribution for Diffusion Models
di: Zhao, Yutian, et al.
Pubblicazione: (2025)

Hessian-informed machine learning interatomic potential towards bridging theory and experiments
di: Yin, Bangchen, et al.
Pubblicazione: (2026)

MD tree: a model-diagnostic tree grown on loss landscape
di: Zhou, Yefan, et al.
Pubblicazione: (2024)

Temporal Graph Analysis with TGX
di: Shirzadkhani, Razieh, et al.
Pubblicazione: (2024)

Unveiling Multi-regime Patterns in SciML: Distinct Failure Modes and Regime-specific Optimization
di: Wang, Yuxin, et al.
Pubblicazione: (2026)

Revisiting inverse Hessian vector products for calculating influence functions
di: Klochkov, Yegor, et al.
Pubblicazione: (2024)

Revisit, Extend, and Enhance Hessian-Free Influence Functions
di: Yang, Ziao, et al.
Pubblicazione: (2024)

Spectral Clustering via Orthogonalization-Free Methods
di: Pang, Qiyuan, et al.
Pubblicazione: (2023)

Policy Bifurcation in Safe Reinforcement Learning
di: Zou, Wenjun, et al.
Pubblicazione: (2024)

A Mathematics Framework of Artificial Shifted Population Risk and Its Further Understanding Related to Consistency Regularization
di: Yang, Xiliang, et al.
Pubblicazione: (2025)

Spectral Condition for $μ$P under Width-Depth Scaling
di: Zheng, Chenyu, et al.
Pubblicazione: (2026)

Graph Generation with Spectral Geodesic Flow Matching
di: Huang, Xikun, et al.
Pubblicazione: (2025)

Meta-Learning Fourier Neural Operators for Hessian Inversion and Enhanced Variational Data Assimilation
di: Moazzami, Hamidreza, et al.
Pubblicazione: (2025)

ADMM-Q: An Improved Hessian-based Weight Quantizer for Post-Training Quantization of Large Language Models
di: Lucas, Ryan, et al.
Pubblicazione: (2026)

Position: Spectral GNNs Are Neither Spectral Nor Superior for Node Classification
di: Jiang, Qin, et al.
Pubblicazione: (2026)

HessFormer: Hessians at Foundation Scale
di: Granziol, Diego
Pubblicazione: (2025)

A Model Zoo on Phase Transitions in Neural Networks
di: Schürholt, Konstantin, et al.
Pubblicazione: (2025)