Copertina

Thermodynamic Irreversibility of Training Algorithms

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Ziyin, Liu, Ren, Yuanjie, Levine, Adam, Chuang, Isaac
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Statistical Mechanics Artificial Intelligence Machine Learning
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Neural Thermodynamics: Entropic Forces in Deep and Universal Representation Learning
di: Ziyin, Liu, et al.
Pubblicazione: (2025)

On Distinguishing Capability Elicitation from Capability Creation in Post-Training: A Free-Energy Perspective
di: Li, Yuhao, et al.
Pubblicazione: (2026)

Trees to Flows and Back: Unifying Decision Trees and Diffusion Models
di: Ramachandran, Sai Niranjan, et al.
Pubblicazione: (2026)

Spontaneous symmetry breaking and Goldstone modes for deep information propagation
di: Iqbal, Nabil, et al.
Pubblicazione: (2026)

Grokking as a Falsifiable Finite-Size Transition
di: Bi, Yuda, et al.
Pubblicazione: (2026)

Reinforcement Learning for Microcanonical Graph Ensemble with Assortativity Constraints
di: Choi, Hoyun, et al.
Pubblicazione: (2026)

Symmetry and Generalisation in Neural Approximations of Renormalisation Transformations
di: Ashworth, Cassidy, et al.
Pubblicazione: (2025)

Cross-fluctuation phase transitions reveal sampling dynamics in diffusion models
di: Ramachandran, Sai Niranjan, et al.
Pubblicazione: (2025)

Stochastic Resetting Mitigates Latent Gradient Bias of SGD from Label Noise
di: Bae, Youngkyoung, et al.
Pubblicazione: (2024)

Thermodynamic Computing via Autonomous Quantum Thermal Machines
di: Lipka-Bartosik, Patryk, et al.
Pubblicazione: (2023)

Maximum diffusion reinforcement learning
di: Berrueta, Thomas A., et al.
Pubblicazione: (2023)

Scaling Autoregressive Models for Lattice Thermodynamics
di: Du, Xiaochen, et al.
Pubblicazione: (2026)

Temporal Memory for Resource-Constrained Agents: Continual Learning via Stochastic Compress-Add-Smooth
di: Chertkov, Michael
Pubblicazione: (2026)

Tensor tree learns hidden relational structures in data to construct generative models
di: Harada, Kenji, et al.
Pubblicazione: (2024)

Sampling Decisions
di: Chertkov, Michael, et al.
Pubblicazione: (2025)

Phase Transitions in the Output Distribution of Large Language Models
di: Arnold, Julian, et al.
Pubblicazione: (2024)

Mixing Artificial and Natural Intelligence: From Statistical Mechanics to AI and Back to Turbulence
di: Chertkov, Michael
Pubblicazione: (2024)

Adaptive Path Integral Diffusion: AdaPID
di: Chertkov, Michael, et al.
Pubblicazione: (2025)

Entropy, concentration, and learning: a statistical mechanics primer
di: Balsubramani, Akshay
Pubblicazione: (2024)

Scalable Discrete Diffusion Samplers: Combinatorial Optimization and Statistical Physics
di: Sanokowski, Sebastian, et al.
Pubblicazione: (2025)

IsingFormer: Augmenting Parallel Tempering With Learned Proposals
di: Bunaiyan, Saleh, et al.
Pubblicazione: (2025)

Relaxation-assisted reverse annealing on nonnegative/binary matrix factorization
di: Haba, Renichiro, et al.
Pubblicazione: (2025)

On the Separability of Information in Diffusion Models
di: Premkumar, Akhil
Pubblicazione: (2025)

Generative Stochastic Optimal Transport: Guided Harmonic Path-Integral Diffusion
di: Chertkov, Michael
Pubblicazione: (2025)

Remove Symmetries to Control Model Expressivity and Improve Optimization
di: Ziyin, Liu, et al.
Pubblicazione: (2024)

Solvation Free Energies from Neural Thermodynamic Integration
di: Máté, Bálint, et al.
Pubblicazione: (2024)

Thermodynamic Performance Limits for Score-Based Diffusion Models
di: Kodama, Nathan X., et al.
Pubblicazione: (2025)

Breaking QAOA's Fixed Target Hamiltonian Barrier: A Fully Connected Quantum Boltzmann Machine via Bilevel Optimization
di: Liu, Jun
Pubblicazione: (2026)

Thermodynamic Isomorphism of Transformers: A Lagrangian Approach to Attention Dynamics
di: Kim, Gunn
Pubblicazione: (2026)

Neural Thermodynamic Integration: Free Energies from Energy-based Diffusion Models
di: Máté, Bálint, et al.
Pubblicazione: (2024)

A Minimal Model of Representation Collapse: Frustration, Stop-Gradient, and Dynamics
di: Yao, Louie Hong, et al.
Pubblicazione: (2026)

Thermodynamic Bayesian Inference
di: Aifer, Maxwell, et al.
Pubblicazione: (2024)

Capturing the Topological Phase Transition and Thermodynamics of the 2D XY Model via Manifold-Aware Score-Based Generative Modeling
di: Jha, Pratyush
Pubblicazione: (2026)

Neural Thermodynamic Laws for Large Language Model Training
di: Liu, Ziming, et al.
Pubblicazione: (2025)

Training thermodynamic computers by gradient descent
di: Whitelam, Stephen
Pubblicazione: (2025)

Analytic Bridge Diffusions for Controlled Path Generation
di: Chertkov, Michael
Pubblicazione: (2026)

Reinforcement Learning for Quantum Technology
di: Bukov, Marin, et al.
Pubblicazione: (2026)

Entropic Confinement and Mode Connectivity in Overparameterized Neural Networks
di: Di Carlo, Luca, et al.
Pubblicazione: (2025)

Support Vector Machine Kernels as Quantum Propagators
di: Kuo, Nan-Hong, et al.
Pubblicazione: (2025)

Privacy-preserving machine learning with tensor networks
di: Pozas-Kerstjens, Alejandro, et al.
Pubblicazione: (2022)