相关论文: Thermodynamic Limits of Physical Intelligence

We study the non-equilibrium thermodynamics of a single particle with two available energy levels, in contact with a classical (Maxwell-Boltzmann) or quantum (Bose-Einstein) heat bath. The particle can undergo transitions between the levels…

The newest and most powerful electronic chips for applications like artificial intelligence generate so much heat that liquid based cooling has become indispensable to prevent breakdown from thermal runaway effects. While cooling schemes…

Maximum entropy has become a mainstream off-policy reinforcement learning (RL) framework for balancing exploitation and exploration. However, two bottlenecks still limit further performance improvement: (1) non-stationary Q-value estimation…

Algorithmic cooling (AC) protocols have been predominantly studied for their cooling capabilities, with limited attention paid to their thermodynamic properties. This work explores a novel perspective by analyzing a broad family of AC…

The minimum heat cost of computation is subject to bounds arising from Landauer's principle. Here, I derive bounds on finite modelling -- the production or anticipation of patterns (time-series data) -- by devices that model the pattern in…

Energy costs of information processing are growing exponentially. Bit erasure is a key problem in this energy-information nexus, and a number of seminal relationships have been deduced regarding the relationship between thermodynamic costs…

Topological insulators are states of matter distinguished by the presence of symmetry protected metallic boundary states. These edge modes have been characterised in terms of transport and spectroscopic measurements, but a thermodynamic…

In this article we want to demonstrate that the time-scale constraints for a thermodynamic system imply the new concept of {\it equipartition of energy bound} (EEB) or, more generally, a thermodynamical bound for the {\it partition} of…

We use a formal correspondence between thermodynamics and inference, where the number of samples can be thought of as the inverse temperature, to study a quantity called ``learning capacity'' which is a measure of the effective…

Performance of the existing physical layer authentication schemes could be severely affected by the imperfect estimates and variations of the communication link attributes used. The commonly adopted static hypothesis testing for physical…

In this paper, we consider the energy cost of communicating using molecular communication. In a simplified scenario, we show that the energy bound in Landauer's principle can be achieved, implying that molecular communication can approach…

Recently, new insights have been obtained by jointly studying communication and resource theory. This interplay consequently serves as a potential platform for interdisciplinary studies. To continue this line, we analyze the role of…

Landauer's principle states that the erasure of one bit of information requires the free energy kT ln 2. We argue that the reliability of the bit erasure process is bounded by the accuracy inherent in the statistical state of the energy…

The dynamic nature of human health and comfort calls for adaptive systems that respond to individual physiological needs in real time. This paper presents an AI-enhanced digital twin framework that integrates biometric signals, specifically…

The metabolic processes complexity is at the heart of energy conversion in living organisms and forms a huge obstacle to develop tractable thermodynamic metabolism models. By raising our analysis to a higher level of abstraction, we develop…

We study the informational underpinnings of thermodynamics and statistical mechanics, using an abstract framework, general probabilistic theories, capable of describing arbitrary physical theories. This allows one to abstract the…

The vision of pervasive artificial intelligence (AI) services can be realized by training an AI model on time using real-time data collected by internet of things (IoT) devices. To this end, IoT devices require offloading their data to an…

Artificial intelligence can outperform humans in specific tasks but consumes substantial energy. How the human brain can work at just 20 watts with complex cognitive intelligence? Here we decode the fundamental information strategy unit of…

Microscopic biological systems operate far from equilibrium, are subject to strong fluctuations, and are composed of many coupled components with interactions varying in nature and strength. Researchers are actively investigating the…

This paper presents an analysis of the fundamental limits on energy efficiency in both digital and analog in-memory computing architectures, and compares their performance to single instruction, single data (scalar) machines specifically in…