惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

WordPress大学
WordPress大学
博客园 - 司徒正美
I
InfoQ
宝玉的分享
宝玉的分享
G
Google Developers Blog
J
Java Code Geeks
Martin Fowler
Martin Fowler
The GitHub Blog
The GitHub Blog
H
Hackread – Cybersecurity News, Data Breaches, AI and More
罗磊的独立博客
腾讯CDC
F
Fortinet All Blogs
A
About on SuperTechFans
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
Recent Announcements
Recent Announcements
Last Week in AI
Last Week in AI
B
Blog RSS Feed
博客园 - 聂微东
D
DataBreaches.Net
Hugging Face - Blog
Hugging Face - Blog
The Cloudflare Blog
L
LangChain Blog
Microsoft Azure Blog
Microsoft Azure Blog
aimingoo的专栏
aimingoo的专栏

cs.IT updates on arXiv.org

Theoretical Limits of Language Model Alignment $f$-Divergence Regularized RLHF: Two Tales of Sampling and Unified Analyses A Unified Measure-Theoretic View of Diffusion, Score-Based, and Flow Matching Generative Models When Can Voting Help, Hurt, or Change Course? Exact Structure of Binary Test-Time Aggregation When Semantic Communication Meets Queueing: Cross-Layer Latency and Task Fidelity Optimization Convexity in Disguise: A Theoretical Framework for Nonconvex Low-Rank Matrix Estimation Conditional Diffusion Under Linear Constraints: Langevin Mixing and Information-Theoretic Guarantees Sharp Capacity Thresholds in Linear Associative Memory: From Winner-Take-All to Listwise Retrieval Expert Routing for Communication-Efficient MoE via Finite Expert Banks Contextual Memory-Enhanced Source Coding for Low-SNR Communications Realizable Bayes-Consistency for General Metric Losses Leveraging Code Automorphisms for Improved Syndrome-Based Neural Decoding A Hierarchical Sampling Framework for bounding the Generalization Error of Federated Learning Dueling DDQN-Based Adaptive Multi-Objective Handover Optimization for LEO Satellite Networks The Causal Description Gap: Information-Theoretic Separations Across Pearl's Hierarchy Optimization of CV-QKD Under Practical Constraints Benchmarking Wireless Representations: High-Dimensional vs. Compressed Embeddings for Efficiency and Robustness Real-Time Text Transmission via LLM-Based Entropy Coding over Fixed-Rate Channels SwiftChannel: Algorithm-Hardware Co-Design for Deep Learning-Based 5G Channel Estimation Evolving Token Communication with Parametric Memory Network Remote Action Generation: Remote Control with Minimal Communication The (Marginal) Value of a Search Ad: An Online Causal Framework for Repeated Second-price Auctions Stabilizing Private LASSO under Heterogeneous Covariates via Anisotropic Objective Perturbation Linear-Readout Floors and Threshold Recovery in Computation in Superposition Soft Graph Diffusion Transformer for MIMO Detection Hierarchical Federated Learning for Networked AI: From Communication Saving to Architecture-Aware Design Exponential families from a single KL identity MIFair: A Mutual-Information Framework for Intersectionality and Multiclass Fairness Diffusion-OAMP for Joint Image Compression and Wireless Transmission Decoupled Descent: Exact Test Error Tracking Via Approximate Message Passing
Thermodynamic limit under Partial Observability
Akihito Sudo · 2021-08-02 · via cs.IT updates on arXiv.org

We study thermodynamic limits when controllers operate with only partial observability of internal correlations in multipartite systems. Understanding the costs imposed by lack of information is crucial in settings where agents must act under incomplete knowledge, and clarifying the thermodynamics of the intertwined sense-act feedback remains a central challenge. We consider a single-step stochastic evolution of a classical multipartite system coupled to a single heat bath under protocols that do not generate new inter-subsystem correlations, and we allow arbitrary reciprocal interaction topology. From Clausius-type local inequalities for each component, we sum and recast the result into an information-theoretic form that isolates the role of inaccessible correlation changes. We thereby obtain a generalized second law: the total entropy production is bounded below exactly by the decrease of internal correlations that is inaccessible to any controller whose state is not included in the relevant information set; only the accessible part can, in principle, be diverted from dissipation. The result refines standard bounds without invoking measurement updates and characterizes information as an indispensable resource: correlations become work-relevant precisely to the extent that they are known, while the unknown portion is unavoidably dissipated.