arXiv subjects

Susanne Still

Publications and source records attributed to Susanne Still.

At least 19 recordsLinked to original sources

arXiv2025

Physical Observers and Quantum Reconstructions

There is a multitude of interpretations of quantum mechanics, but foundational principles are lacking. Relational quantum mechanics views the observer as a physical system, which allows for an unambiguous interpretation as all axioms are purely operational, describing how observers acquire information. The approach, however, is based on the premise that the observer retains only predictive information about the observed system. Here, we justify this premise using the following principle: Physically embedded observers choose information processing strategies that provide them with the option to approach physical limits to the greatest possible extent. Applied to a lower limit on energy dissipation, the principle leads directly to a compact predictive model, thus justifying this core premise of relational quantum mechanics.

quant-ph↗

arXiv2023

Thermodynamically rational decision making under uncertainty

An analytical characterization of thermodynamically rational agent behaviour is obtained for a simple, yet non--trivial example of a ``Maxwell's demon" operating with partial information. Our results provide the first fully transparent physical understanding of a decision problem under uncertainty.

physics.data-an↗

arXiv2023

The physical observer in a Szilard engine with uncertainty

Information engines model ``Maxwell's demon" mechanistically. However, the demon's strategy is pre-described by an external experimenter, and information engines are conveniently designed such that observables contain complete information about variables pertinent to work extraction. In real world scenarios, it is more realistic to encounter partial observability, which forces the physical observer, an integral part of the information engine, to make inferences from incomplete knowledge. Here, we use the fact that an algorithm for computing optimal strategies can be directly derived from maximizing overall engine work output. For a simple binary decision problem, we discover interesting optimal strategies that differ notably from naive coarse graining. They inspire a model class of simple, yet compelling, parameterized soft partitionings of the observable.

cond-mat.stat-mech↗

arXiv2022

Information engine in a nonequilibrium bath

Information engines can convert thermal fluctuations of a bath at temperature $T$ into work at rates of order $k_\mathrm{B}T$ per relaxation time of the system. We show experimentally that such engines, when in contact with a bath that is out of equilibrium, can extract much more work. We place a heavy, micron-scale bead in a harmonic potential that ratchets up to capture favorable fluctuations. Adding a fluctuating electric field increases work extraction up to ten times, limited only by the strength of applied field. Our results connect Maxwell's demon with energy harvesting and an estimate of efficiency shows that information engines in nonequilibrium baths can greatly outperform conventional engines.

cond-mat.stat-mech↗

arXiv2022

Energetic cost of feedback control

Successful feedback control of small systems allows for the rectification of thermal fluctuations, converting them into useful energy; however, control itself requires work. This paper emphasizes the fact that the controller is a physical entity interacting with the feedback-controlled system. For a specifically designed class of controllers, reciprocal interactions become nonreciprocal due to large timescale separation, which considerably simplifies the situation. We introduce a minimally dissipative controller model, illustrating the findings using a simple example. We find that the work required to run the controller must at least compensate for the decrease in entropy due to the control operation.

cond-mat.stat-mech↗

arXiv2021

The Evolution of Imitation Without Cultural Transmission

The evolution and function of imitation have always been placed within the confines of animal learning and associated with its crucial role in cultural transmission and cultural evolution. Can imitation evolve as a form of phenotypic plasticity in the absence of cultural transmission, in phenotypes beyond behavior? We investigate a model in which imitation is unbundled from cultural transmission: an organism's adult phenotype is plastically altered by its experiences as a juvenile of other juveniles' genetically determined traits. The only information transmitted between generations is genetic. We find that during a period of directional selection towards a phenotypic optimum, natural selection favors modifiers which cause an organism to bias its plastic phenotype in the direction opposite to the mean phenotype of the population -- anti-imitation. As the population approaches the phenotypic optimum and shifts into stabilizing selection, selection on the modifier reverses and favors strong imitation of the population mean. Imitation can evolve to overshoot the target and produce an evolutionary pathology where mean fitness decreases. When purifying selection for an extreme phenotype is modeled, only selection for anti-imitation occurs, even at a mutation-selection balance. Imitation and anti-imitation emerge from these models in the absence of any goal representation, cognitive understanding of its purpose, or discernment of any kind. These theoretical outcomes are all novel evolutionary and biological phenomena, and we discuss their implications.

q-bio.PE↗

arXiv2021

Optimal work extraction and mutual information in a generalized Szilárd engine

A 1929 Gedankenexperiment proposed by Szilárd, often referred to as "Szilárd's engine", has served as a foundation for computing fundamental thermodynamic bounds to information processing. While Szilárd's original box could be partitioned into two halves and contains one gas molecule, we calculate here the maximal average work that can be extracted in a system with $N$ particles and $q$ partitions, given an observer which counts the molecules in each partition, and given a work extraction mechanism that is limited to pressure equalization. We find that the average extracted work is proportional to the mutual information between the one-particle position and the vector containing the counts of how many particles are in each partition. We optimize this quantity over the initial locations of the dividing walls, and find that there exists a critical number of particles $N^{\star}(q)$ below which the extracted work is maximized by a symmetric configuration of the $q$ partitions, and above which the optimal partitioning is asymmetric. Overall, the average extracted work is maximized for a number of particles $\hat{N}(q)<N^{\star}(q)$, with a symmetric partition. We calculate asymptotic values for $N\rightarrow \infty$.

cond-mat.stat-mech↗

arXiv2021

Partially Observable Szilard Engines

Leo Szilard pointed out that Maxwell's demon can be replaced by machinery, thereby laying the foundation for understanding the physical nature of information. Szilard's information engine still serves as a canonical example after almost a hundred years, despite recent significant growth of the area. The role the demon plays can be reduced to mapping observable data to a meta-stable memory, which is utilized to extract work. While Szilard showed that the map can be implemented mechanistically, it was chosen a priori. The choice of how to construct a meaningful memory constitutes the demon's intelligence. Recently, it was shown that this can be automated as well. To that end, generalized, partially observable information engines were introduced, providing a basis for understanding the physical nature of information processing. Partial observability is ubiquitous in real world systems which have limited sensor types and information acquisition bandwidths. Generalized information engines can run work extraction at a different temperature, T' > T, from the memory forming process. This enables the combined treatment of heat engines and information engines. We study the physical characteristics of intelligent observers by introducing a canonical model that displays physical richness, despite its simplicity. A minor change to Szilard's engine - inserting the divider at an angle - results in a family of partially observable Szilard engines. Their analysis shows how the demon's intelligence can be automated. For each angle, and for each value of T'/T, an optimal memory can be found, enabling the engine to run with minimal dissipation. Those optimal memories are probabilistic maps, computed algorithmically. We discuss how they can be implemented with a simple physical system, characterize their performance, and compare their quality to that of naive, deterministic quantizations of the observable.

cond-mat.stat-mech↗

arXiv2019

Thermodynamic cost and benefit of memory

This letter exposes a tight connection between the thermodynamic efficiency of information processing and predictive inference. A generalized lower bound on dissipation is derived for partially observable information engines which are allowed to use temperature differences. It is shown that the retention of irrelevant information limits efficiency. A data representation strategy is derived from optimizing a fundamental physical limit to information processing: minimizing the lower bound on dissipation leads to a data compression method that maximally retains relevant, predictive, information. In that sense, predictive inference emerges as the strategy that least precludes energy efficiency.

cond-mat.stat-mech↗

arXiv2018

Physical Limitations of Work Extraction from Temporal Correlations

Recently proposed information-exploiting systems designed to extract work from a single heat bath utilize temporal correlations on an input tape. We study how enforcing time-continuous dynamics, which is necessary to ensure the device is physically realizable, constrains possible designs and drastically diminishes efficiency. We show that these problems can be circumvented by means of applying an external, time-varying protocol. This turns the device from a "passive", free-running machine into an "actively" driven one.

cond-mat.stat-mech↗

arXiv2017

Non-equilibrium odds for the emergence of life

Large and complex molecules are building blocks for life. We compute probabilities for their formation from an average non-equilibrium model. As the distance from thermodynamic equilibrium is increased in this model, so too are the chances for forming molecules that would be prohibitively rare in thermodynamic equilibrium. This effect is explored in two settings: the synthesis of heavy amino acids, and their polymerization into peptides. In the extreme non-equilibrium limit, concentrations of the heaviest amino acids can be boosted by a factor of 10,000. Concentrations of the longest peptide chains can be increased by hundreds of orders of magnitude. Since all details of the non-equilibrium driving are averaged out, these findings indicate that, independent of the details of the driving, the mere fact that pre-biotic environments were not in thermodynamic equilibrium may help cross the barriers to the formation of life.

physics.bio-ph↗

arXiv2015

On Improving the Performance of Nonphotochemical Quenching in CP29 Light-Harvesting Antenna Complex

We model and simulate the performance of charge-transfer in nonphotochemical quenching (NPQ) in the CP29 light-harvesting antenna-complex associated with photosystem II (PSII). The model consists of five discrete excitonic energy states and two sinks, responsible for the potentially damaging processes and charge-transfer channels, respectively. We demonstrate that by varying (i) the parameters of the chlorophyll-based dimer, (ii) the resonant properties of the protein-solvent environment interaction, and (iii) the energy transfer rates to the sinks, one can significantly improve the performance of the NPQ. Our analysis suggests strategies for improving the performance of the NPQ in response to environmental changes, and may stimulate experimental verification.

physics.bio-ph↗

arXiv2015

Quantum Predictive Filtering

How can relevant information be extracted from a quantum process? In many situations, only some part of the total information content produced by an information source is useful. Can one then find an efficient encoding, in the sense of retaining the largest fraction of relevant information? This paper offers one possible solution by giving a generalization of a classical method designed to retain as much relevant information as possible in a lossy data compression. A key feature of the method is to introduce a second information source to define relevance. We quantify the advantage a quantum encoding has over the best classical encoding in general, and we demonstrate using examples that a substantial quantum advantage is possible. A main result, however, is that if the relevant information is purely classical, then a classical encoding is optimal.

quant-ph↗

arXiv2014

$L_p$ regularized portfolio optimization

Investors who optimize their portfolios under any of the coherent risk measures are naturally led to regularized portfolio optimization when they take into account the impact their trades make on the market. We show here that the impact function determines which regularizer is used. We also show that any regularizer based on the norm $L_p$ with $p>1$ makes the sensitivity of coherent risk measures to estimation error disappear, while regularizers with $p<1$ do not. The $L_1$ norm represents a border case: its "soft" implementation does not remove the instability, but rather shifts its locus, whereas its "hard" implementation (equivalent to a ban on short selling) eliminates it. We demonstrate these effects on the important special case of Expected Shortfall (ES) that is on its way to becoming the next global regulatory market risk measure.

q-fin.PM↗

arXiv2012

The thermodynamics of prediction

A system responding to a stochastic driving signal can be interpreted as computing, by means of its dynamics, an implicit model of the environmental variables. The system's state retains information about past environmental fluctuations, and a fraction of this information is predictive of future ones. The remaining nonpredictive information reflects model complexity that does not improve predictive power, and thus represents the ineffectiveness of the model. We expose the fundamental equivalence between this model inefficiency and thermodynamic inefficiency, measured by dissipation. Our results hold arbitrarily far from thermodynamic equilibrium and are applicable to a wide range of systems, including biomolecular machines. They highlight a profound connection between the effective use of information and efficient thermodynamic operation: any system constructed to keep memory about its environment and to operate with maximal energetic efficiency has to be predictive.

cond-mat.stat-mech↗

arXiv2011

Optimal Liquidation Strategies Regularize Portfolio Selection

We consider the problem of portfolio optimization in the presence of market impact, and derive optimal liquidation strategies. We discuss in detail the problem of finding the optimal portfolio under Expected Shortfall (ES) in the case of linear market impact. We show that, once market impact is taken into account, a regularized version of the usual optimization problem naturally emerges. We characterize the typical behavior of the optimal liquidation strategies, in the limit of large portfolio sizes, and show how the market impact removes the instability of ES in this context.

q-fin.PM↗

arXiv2010

Optimal Causal Inference: Estimating Stored Information and Approximating Causal Architecture

We introduce an approach to inferring the causal architecture of stochastic dynamical systems that extends rate distortion theory to use causal shielding---a natural principle of learning. We study two distinct cases of causal inference: optimal causal filtering and optimal causal estimation. Filtering corresponds to the ideal case in which the probability distribution of measurement sequences is known, giving a principled method to approximate a system's causal structure at a desired level of representation. We show that, in the limit in which a model complexity constraint is relaxed, filtering finds the exact causal architecture of a stochastic dynamical system, known as the causal-state partition. From this, one can estimate the amount of historical information the process stores. More generally, causal filtering finds a graded model-complexity hierarchy of approximations to the causal architecture. Abrupt changes in the hierarchy, as a function of approximation, capture distinct scales of structural organization. For nonideal cases with finite data, we show how the correct number of underlying causal states can be found by optimal causal estimation. A previously derived model complexity control term allows us to correct for the effect of statistical fluctuations in probability estimates and thereby avoid over-fitting.

cs.IT↗

arXiv2009

Regularizing Portfolio Optimization

The optimization of large portfolios displays an inherent instability to estimation error. This poses a fundamental problem, because solutions that are not stable under sample fluctuations may look optimal for a given sample, but are, in effect, very far from optimal with respect to the average risk. In this paper, we approach the problem from the point of view of statistical learning theory. The occurrence of the instability is intimately related to over-fitting which can be avoided using known regularization methods. We show how regularized portfolio optimization with the expected shortfall as a risk measure is related to support vector regression. The budget constraint dictates a modification. We present the resulting optimization problem and discuss the solution. The L2 norm of the weight vector is used as a regularizer, which corresponds to a diversification "pressure". This means that diversification, besides counteracting downward fluctuations in some assets by upward fluctuations in others, is also crucial because it improves the stability of the solution. The approach we provide here allows for the simultaneous treatment of optimization and diversification in one framework that enables the investor to trade-off between the two, depending on the size of the available data set.

q-fin.PM↗