Date25, Sep., 2023, (Mon.), 14:00~
PlaceRoom 413 in Faculty of Science Building No.1, The University of Tokyo
SpeakerAmit Kumar Chatterjee (Kyoto University)
TitleQuantum Mpemba effect: an anomalous relaxation in quantum systems
AbstractMpemba effect refers to the counter-intuitive phenomenon where a hotter object can cool down faster than a colder copy of the same object. In spite of some theoretical as well as experimental advances in the classical domain, the quantum counterpart of the Mpemba effect, specifically in temperature, has remained unexplored. In this talk, we demonstrate the quantum Mpemba effect by showing that temperatures of two copies of a quantum system, one initially hotter than the other, can cross each other after some time and thereafter reverse their identities, i.e. hotter becomes colder and vice versa, before reaching the same final temperature. In fact, we show such crossing of trajectories characterizing the quantum Mpemba effect, can occur in several other observables including energy, entropy, distance function etc. Our theoretical results on quantum Mpemba effect are primarily based on a quantum dot connected to two reservoirs. In the later part of the talk, we discuss how exceptional points and complex eigenvalues can lead to multiple quantum Mpemba effect (where trajectories cross multiple times) in a two-level driven dissipative system.
A. K. Chatterjee, S. Takada, and H. Hayakawa, Phys. Rev. Lett. 131, 080402 (2023).
Date2, Aug., 2023, (Wed.), 14:00~
PlaceRoom 413 in Faculty of Science Building No.1, The University of Tokyo / Zoom (If you would like to join, please send an email to sosuke.ito(at)ubi.s.u-tokyo.ac.jp.)
SpeakerMiguel Aguilera, (Basque Center for Applied Mathematics)
TitleNonequilibrium Neural Computation: Stochastic thermodynamics of the asymmetric Sherrington-Kirkpatrick model
AbstractMost systems in nature operate far from equilibrium, exhibiting time-asymmetric, irreversible dynamics; giving rise to entropy production as they exchange energy and matter with their environment. In neuroscience, effective information processing entails flexible architectures integrating multiple sensory streams that vary in time with internal and external events. Physically, neural computation is, in a thermodynamic sense, an out-of-equilibrium, non-stationary process that changes dynamically. Cognitively, nonequilibrium neural activity results in dynamic changes in sensory streams and internal states. In contrast, classical neuroscience theory focuses on stationary, equilibrium information paradigms (e.g., efficient coding theory), which often fail to describe the role of nonequilibrium fluctuations in neural processes.
Inspired by the success of the equilibrium Ising model in investigating disordered systems and related associative-memory neural networks, we study the nonequilibrium thermodynamics of the asymmetric Sherrington-Kirkpatrick system as a prototypical model of large-scale nonequilibrium networks. We employ a path integral method to calculate a generating functional over the trajectories to derive exact solutions of the order parameters, conditional entropy of trajectories, and steady-state entropy production of infinitely large networks. We find that entropy production peaks at a critical order-disorder phase transition but is more prominent in a regime with quasi-deterministic disordered dynamics. While entropy production is becoming popular to characterize various complex systems as well as neural activity, our results reveal that increased entropy production is linked with radically different scenarios, and combining multiple thermodynamic quantities yields a more precise picture of the system. These results contribute to an exact analytical theory for studying the thermodynamic properties of large-scale nonequilibrium systems and their phase transitions. 

Refs: Aguilera, M., Igarashi, M. & Shimazaki, H. Nonequilibrium thermodynamics of the asymmetric Sherrington-Kirkpatrick model. Nature Communications 14, 3685 (2023). https://doi.org/10.1038/s41467-023-39107-y
Date18, Jul., 2023, (Tue.), 14:00~
PlaceRoom 413 in Faculty of Science Building No.1, The University of Tokyo
SpeakerJean-Charles Delvenne, (UCLouvain)
Title (Non-)stochastic thermodynamics for computing devices 
AbstractIn this talk I describe several theoretical bounds and illustrate them on how they can place trade-offs on the performance (in terms of dissipation, reliability and speed) of electronic devices and circuits made from them.  These results are consistent with numerical simulations and experimental measurements.  
Date30, Nov., 2022, (Wed.), 14:00~
PlaceRoom 413 in Faculty of Science Building No.1, The University of Tokyo / Zoom (If you would like to join, please send an email to sosuke.ito(at)ubi.s.u-tokyo.ac.jp.)
Speaker竹内尚輝氏 (横浜国立大学)
Title断熱超伝導回路 —基礎と可逆計算機への応用—
Date27, Oct., 2022, (Thu.), 15:00~
PlaceRoom 413 in Faculty of Science Building No.1, The University of Tokyo / Zoom (If you would like to join, please send an email to sosuke.ito(at)ubi.s.u-tokyo.ac.jp.)
SpeakerNishide Ryosuke (the University of Tokyo) 
TitlePattern propagation driven by surface curvature
AbstractPattern formation occurs on curved surfaces abundantly especially in biological systems. Previous studies have revealed that the surface curvature affects the pattern dynamics and plays biological functions, however, comprehensive understanding is still elusive. Here by employing reaction-diffusion systems showing Turing pattern, we show for the first time that static pattern on a flat surface can be propagating wave on a curved surface. By numerical and theoretical analyses, it is shown that the pattern propagation is conditioned by the symmetry of surface and pattern. We also show the results of weakly nonlinear analysis applied to the problem, which suggests rich dynamics can arise on curved surface.
Date30, May., 2022, (Mon.), 13:30~
PlaceRoom 413 in Faculty of Science Building No.1, The University of Tokyo
SpeakerYoshiya Matsubara (National Centre for Biological Sciences) 
TitleError catastrophe can be avoided by proofreading innate to template-directed polymerization
AbstractHow and if information is maintained through polymer replication is among the most fundamental issues in studies of the origins of life, as first noted by Manfred Eigen. The “error catastrophe” problem in replication asserts that maintenance of information is more difficult for longer polymers due to thermodynamically inevitable errors. In this study, we analyzed the population dynamics of replicating templates explicitly incorporated with the kinetics of the fundamental polymerization process. Numerical and theoretical analyses suggest that the template-directed polymerization process entails an inherent error-correction mechanism akin to the kinetic proofreading proposed by J. J. Hopfield. Interestingly, because of such effects, the tolerance to errors increases with the length of the replicating template polymer, which solves the error-catastrophe problem. Therefore, the findings provide novel principles for error correction without sophisticated and specific mechanisms that are potentially applicable to replication under origins of life scenarios.

Y. J. Matsubara, N. Takeuchi, & K. Kaneko arXiv preprint arXiv:2108.09961 (2021) 
Date27, May., 2022, (Fri.), 16:00~17:00
PlaceZoom (If you would like to join, please send an email to sosuke.ito(at)ubi.s.u-tokyo.ac.jp.)
SpeakerMauricio del Razo Sarmina (Free University of Berlin) 
TitleChemical diffusion master equation: stochastic formulations of reaction-diffusion processes
AbstractBiological processes occurring at scales of a biological cell can be described in terms of diffusing and interacting molecules. At the level of particle-resolved descriptions, where molecules are represented by particles and chemical reactions are coupled to their spatial diffusion, there exist comprehensive numerical simulation schemes, while the corresponding mathematical formalization is relatively underdeveloped. In this work, we provide different frameworks to systematically formulate the probabilistic evolution equation, termed chemical diffusion master equation (CDME), that governs stochastic reaction-diffusion processes. Such frameworks can be used to unify reaction-diffusion models at different scales, allowing the development of consistent multiscale simulation schemes. Moreover, these frameworks enable theories, such as stochastic thermodynamics, to be applied to reaction-diffusion processes. This can result in fundamental insights of the physical capabilities and limitations of biological systems and beyond.
Date8, Dec., 2021(Wed.) 10:00~
PlaceZoom (If you would like to join, please send an email to sosuke.ito(at)ubi.s.u-tokyo.ac.jp.)
SpeakerArtemy Kolchinsky (Santa Fe Institute)
TitleThermodynamic threshold for Darwinian evolution
AbstractUnderstanding the thermodynamics of Darwinian evolution has important implications for biophysics, evolutionary biology, and the study of the origin of life. We show that in a population of nonequilibrium autocatalytic replicators, the critical selection coefficient (i.e., the minimal fitness difference visible to selection) is lower bounded by the free energy dissipated per replication event. This bound represents a fundamental thermodynamic threshold for Darwinian evolution, analogous to selection thresholds that arise from finite population sizes or large mutation rates. Our results apply to a large class of molecular replicators, including many types of multistep autocatalytic reaction mechanisms and autocatalytic sets. We illustrate our approach on a thermodynamically-consistent model of simple replicators in a chemostat.
Date2021年4月20日(火) 14:00~
PlaceZoom (参加をご希望の方は sosuke.ito(at)ubi.s.u-tokyo.ac.jp までご連絡ください.)
Speaker磯村拓哉氏 (理化学研究所脳神経科学研究センター)
Abstract神経科学には2つの主要な数理モデリングアプローチが存在する。1つは単純な微分方程式を用いた力学系アプローチであり、リザーバーネットワークなどの形式により、神経回路のダイナミクスをもっともらしく解釈することができる(Sussillo & Abbott, Neuron, 2009)。もう一つは、自由エネルギー原理や能動的推論(active inference)で考えられているように、脳を変分ベイズ推論を行うエージェントとみなす考え方である(Friston, Nat Rev Neurosci, 2010)。しかし、この2つの主要なアプローチの対応関係は完全には理解されていない。そこで本研究では、神経活動と可塑性が共通のコスト関数を最小化しているような、発火率コーディングモデルからなる標準的な神経回路モデルのクラスを考えた。我々は、これらの神経回路モデルが変分ベイズ推論を暗に実行していること数理的に示した(Isomura & Friston, Neural Comput, 2020)。すなわち、この種の神経ダイナミクスは、部分観測マルコフ決定過程モデルの下での変分自由エネルギー最小化を行なっているとみなすことができる。この等価性を用いると、ベイズ推論の観点から標準的な神経回路モデルの普遍的な特性を説明することができ、適応的行動制御や行動計画の神経メカニズムを理解する助けになると期待される。
Date2020年10月6日(火) 13:30~
PlaceZoom (参加をご希望の方は sosuke.ito(at)ubi.s.u-tokyo.ac.jp までご連絡ください.)
Speaker日浦健氏 (京都大学)
Abstract平衡統計力学が熱力学第二法則と整合的であることはよく知られている.逆に,第二法則,特に第二種永久機関が存在しないことを前提にしたとき,等重率の原理のような統計力学の確率論的な記述はどのように理解できるだろうか.この一見とらえどころのない問いをある仕方で形式化すると,Shafer と Vovk らのゲーム論的確率論と全く同一の形式で議論できることがわかる.彼らの確率論の特徴は,測度を用いることなく,何人かのプレイヤー間のギャンブルを通して大数の法則のような統計法則や確率・期待値といった概念を定式化することにある.本講演では,繰り返しサイクル操作による仕事の取り出しを,仕事を取り出そうとする「我々」と第二法則を保持しながら状態を出力する「自然」の間のギャンブルとして定式化し,「我々」がある戦略を採用することによって,ギブス分布が経験分布の意味で実現されることを示す.また,「我々」に可能な操作が少数の外場のみ制御できるような制限された場合であっても,外場に共役な変数の経験平均を平衡値に収束させるような戦略が存在することを示す.ゲーム論的確率論の入門や背景にあるマルチンゲール理論についても簡単に解説する予定である.
Date16, Jun., 2020, (Tue.), 15:00~
PlaceZoom (If you would like to join, please send an email to sosuke.ito(at)ubi.s.u-tokyo.ac.jp.)
SpeakerTan Van Vu (The University of Tokyo)
TitleGeometrical bounds of the irreversibility in Markovian systems
AbstractWe derive geometrical bounds on the irreversibility for both classical and open quantum Markovian systems that satisfy the detailed balance conditions. Using the information geometry, we prove that the irreversible entropy production is bounded from below by a modified Wasserstein distance between the initial and final states, thus generalizing the Clausius inequality. The modified Wasserstein metric can be regarded as a discrete-state generalization of the Wasserstein metric, which plays an important role in the optimal transport theory. Notably, the derived bounds can be interpreted as classical and quantum speed limits, implying that the associated entropy production constrains the minimum time required to transform a system state. We illustrate the results on several systems and demonstrate that a tighter bound on the efficiency of quantum heat engines can be obtained. 
Date23, Jan., 2020, (Thu.), 15:00~
PlaceRoom 413 in Faculty of Science Building No.1, The University of Tokyo
SpeakerSeth Fraden (Physics, Brandeise University)
TitleProgrammable self-assembly of DNA origami capsids based on the principles of virus structure
AbstractWe provide a general and modular solution for building synthetic icosahedral shells on the scale of 100 nm, motivated by the 1962 Caspar and Klug theory of virus structure. Strategies were explored for controlling the pathways, kinetics, and the yield by which subunits arrange themselves into icosahedral symmetry. The methods of DNA origami were employed to produce accurately-designed and rigid building blocks. We created multiple large virus-like capsids and validated the structures using cryo electron microscopy and studied the capsid assembly process experimentally and with a computational model to elucidate how the kinetics and yield of target structures depends on control parameters. Our capsid building blocks represent a near-ideal manifestation of patchy particles whose geometry and interactions can be designed with sub-nanometer and kBT precision, thus achieving a long sought after goal in soft matter physics. Applications range from drug delivery to a generalized antiviral agent, which is demonstrated for hepatitis B. 
Date21, Jan., 2020, (Tue.), 13:00~
PlaceRoom 413 in Faculty of Science Building No.1, The University of Tokyo
SpeakerYan Jiawei (Harvard Medical School)
TitleKinetic uncertainty relations for the control of stochastic reaction systems
AbstractNon-equilibrium stochastic reaction networks are commonly found in both biological and non-biological systems, but have remained hard to analyze because small differences in rate functions or topology can change the dynamics drastically. Here we conjecture exact quantitative inequalities that relate the extent of fluctuations in connected components, for various network topologies. Specifically, we find that regardless of how two components affect each other’s production rates, it is impossible to suppress fluctuations below the uncontrolled equivalents for both components: one must increase its fluctuations for the other to be suppressed. For systems in which components control each other in ring-like structures, it appears that fluctuations can only be suppressed in one component if all other components instead increase fluctuations, compared to the case without control. Even the general N-component system, with arbitrary connections and parameters, must have at least one component with increased fluctuations to reduce fluctuations in others. In connected reaction networks it thus appears impossible to reduce the statistical uncertainty in all components, regardless of the control mechanisms or energy dissipation. 
Date9, Dec., 2019, (Thu.), 15:00~
PlaceRoom 1320 in Faculty of Science Building No. 4, The University of Tokyo
SpeakerDavid Wolpert (Santa Fe Institute)
TitleThe stochastic thermodynamics of computation
AbstractThis seminar is mainly organized by Sagawa Lab.. One of the central concerns of computer science is how the resources needed to perform a given computation depend on that computation. Moreover, one of the major resource requirements of computers—ranging from biological cells to human brains to high-performance (engineered) computers—is the energy used to run them, i.e. the thermodynamic costs of running them. Those thermodynamic costs of performing a computation have been a long-standing focus of research in physics, going back (at least) to the early work of Landauer and colleagues. However, one of the most prominent aspects of computers is that they are inherently non-equilibrium systems. Unfortunately, the research by Landauer and co-workers on the thermodynamics of computation was done when non-equilibrium statistical physics was still in its infancy, severely limiting the scope and formal detail of their analyses. The recent breakthroughs in non-equilibrium statistical physics hold the promise of allowing us to go beyond those limitations. Here I present some initial results along these lines, concerning the entropic costs of running (loop-free) digital circuits and Turing machines. These results reveal new, challenging engineering problems for how to design computers to have minimal thermodynamic costs. They also allow us to start to combine computer science theory and stochastic thermodynamics at a foundational level, thereby expanding both. 
Date31, Oct, 2019, (Thu.), 14:00~
PlaceRoom 413 in Faculty of Science Building No. 1, The University of Tokyo
SpeakerSreekanth K. Manikandan (Stockholm University)
TitleInferring entropy production from short experiments
AbstractThis seminar is mainly organized by Sagawa Lab.. We provide a strategy for an exact inference of the average as well as the fluctuations of the entropy production in non-equilibrium systems in the steady state, from the measurements of arbitrary current fluctuations. Our results are built upon the finite time generalization of the thermodynamic uncertainty relation, and require only very short time series data from experiments. We illustrate our results with exact and numerical solutions for two colloidal heat engines.Arxiv link: https://arxiv.org/abs/1910.00476 
Date2019年9月2日(月) 15:00~
Place東京大学 理学部1号館 4階413号室
TitleLévy flightの統計力学:生物物理系におけるミクロモデルからの導出
Abstract 統計力学において拡散現象は重要なトピックである.最も素朴な拡散現象は「通常拡散」であり,典型的にはブラウン運動で記述され,その変位はガウス分布に従う.ブラウン運動を統計力学として理解することは古くから試みられてきたことであり,例えばニュートン力学を出発点に,ボルツマン方程式,ランジュバン方程式を漸近的に導く運動論に基づく数理的な枠組み[1]や,射影演算子法に基づく導出[2]などがよく知られている.一方,通常拡散の枠組みに当てはまらない「異常拡散」と呼ばれる拡散現象も存在する.異常拡散の典型例の一つはLévy flightと呼ばれるモデルであり,変位幅が冪分布に従うため,間欠的に大きなジャンプを示す数理モデルとなっている.Lévy flightは非平衡のモデルであり,乱流,生物,経済などの幅広い分野で観測され大きな着目を浴びてきた.それではLévy flightを統計力学として導出するにはどうすればよいだろうか?講演者の調べた限り,ミクロモデルの動力学から出発して体系的にLévy flightを導出する試みは殆どない.Lévy flightは本質的に非平衡系のモデルであり,非平衡状態の多粒子系動力学を出発点に取ることは理論的に容易でないからだ.そこで講演者は生物物理の非平衡系に対して運動論の枠組みを拡張することで,Lévy flightをミクロモデルから体系的に導出する研究を行った[3].講演者は流体中の遊走微生物系をモデルケースとして研究した.この設定では遊走微生物が流体力学相互作用を引き起こすことで,非熱的な揺らぎを流体に伝播させる.ここで水中にトレーサ粒子を入れると,流体を伝播した非熱的な揺らぎによってトレーサが変位を示す.講演者のグループはこのトレーサの変位がLévy flightに従うことを,そのミクロ動力学を出発点に統計力学として示した[3].ここでの基本的なアイディアは分子運動論の数理を拡張することである.遊走微生物の分布が希薄になる極限では,トレーサの動きはカラード・ポアソンモデルとして近似的に記述できる.ここでのノイズ中に含まれる力の関数や,ノイズの発生頻度は,非平衡2体散乱過程の力学を解析的に解くことによって求める.このカラード・ポアソンモデルを解析することで,長時間の振る舞いとしてLévy flightが現れることを示した.またこの解析の副産物として,Holtsmark型の静的分布関数論に基づく現象論的アプローチが短時間極限で正当化されることも示された[4].[1] N.G. van Kampen, Stochastic Processes in Physics and Chemistry, 3rd ed. (North-Holland, 2007)[2] 川崎 恭治,非平衡と相転移―メソスケールの統計物理学 (朝倉書店,2000)[3] K. Kanazawa, T.G. Sano, A. Cairoli, and A. Baule: arXiv:1906.00608 (2019)[4] T. Kurihara et al. Phys. Rev. E 95, 030601 (2017)
Place東京大学 本郷キャンパス
WorkshopData analysis and machine learning in dynamical systems (website)
AbstractThe goal of this workshop is to bring together researchers from data analysis, machine learning, and dynamical systems to discuss recent progress in data analysis of complex phenomena in dynamical systems with large degrees of freedom, and to fill the gap between theories in these fields. 
Date2019年2月8日(金) 16:00~
Place東京大学 理学部1号館 4階413号室