arXiv subjects

Shreyas Gokhale

Publications and source records attributed to Shreyas Gokhale.

15 recordsLinked to original sources

arXiv2026

Topological flowscape reveals state transitions in nonreciprocal living matter

Nonreciprocal interactions -- where forces between entities are asymmetric -- govern a wide range of nonequilibrium phenomena, yet their role in structural transitions in living and active systems remains elusive. Here, we demonstrate a transition between nonreciprocal states using starfish embryos at different stages of development, where interactions are inherently asymmetric and tunable. Experiments, interaction inference, and topological analysis yield a nonreciprocal state diagram spanning crystalline, flocking, and fragmented states, revealing that weak nonreciprocity promotes structural order while stronger asymmetry disrupts it. To capture these transitions, we introduce topological landscapes, mapping the distribution of structural motifs across state space. We further develop topological flowscapes, a dynamic framework that quantifies transitions between collective states and detects an informational rate shift from the experimental state transition. Together, these results establish a general approach for decoding nonequilibrium transitions and uncover how asymmetric interactions sculpt the dynamical and structural architecture of active and living matter.

cond-mat.soft↗

arXiv2025

MonoCon: A general framework for learning ultra-compact high-fidelity representations using monotonicity constraints

Learning high-quality, robust, efficient, and disentangled representations is a central challenge in artificial intelligence (AI). Deep metric learning frameworks tackle this challenge primarily using architectural and optimization constraints. Here, we introduce a third approach that instead relies on $\textit{functional}$ constraints. Specifically, we present MonoCon, a simple framework that uses a small monotonic multi-layer perceptron (MLP) head attached to any pre-trained encoder. Due to co-adaptation between encoder and head guided by contrastive loss and monotonicity constraints, MonoCon learns robust, disentangled, and highly compact embeddings at a practically negligible performance cost. On the CIFAR-100 image classification task, MonoCon yields representations that are nearly 9x more compact and 1.5x more robust than the fine-tuned encoder baseline, while retaining 99\% of the baseline's 5-NN classification accuracy. We also report a 3.4x more compact and 1.4x more robust representation on an SNLI sentence similarity task for a marginal reduction in the STSb score, establishing MonoCon as a general domain-agnostic framework. Crucially, these robust, ultra-compact representations learned via functional constraints offer a unified solution to critical challenges in disparate contexts ranging from edge computing to cloud-scale retrieval.

cs.LG↗

arXiv2024

Selective excitation of work-generating cycles in nonreciprocal living solids

Emergent nonreciprocity in active matter drives the formation of self-organized states that transcend the behaviors of equilibrium systems. Integrating experiments, theory and simulations, we demonstrate that active solids composed of living starfish embryos spontaneously transition between stable fluctuating and oscillatory steady states. The nonequilibrium steady states arise from two distinct chiral symmetry breaking mechanisms at the microscopic scale: the spinning of individual embryos resulting in a macroscopic odd elastic response, and the precession of their rotation axis, leading to active gyroelasticity. In the oscillatory state, we observe long-wavelength optical vibrational modes that can be excited through mechanical perturbations. Strikingly, these excitable nonreciprocal solids exhibit nonequilibrium work generation without cycling protocols, due to coupled vibrational modes. Our work introduces a novel class of tunable nonequilibrium processes, offering a framework for designing and controlling soft robotic swarms and adaptive active materials, while opening new possibilities for harnessing nonreciprocal interactions in engineered systems.

cond-mat.soft↗

arXiv2023

The semantic landscape paradigm for neural networks

Deep neural networks exhibit a fascinating spectrum of phenomena ranging from predictable scaling laws to the unpredictable emergence of new capabilities as a function of training time, dataset size and network size. Analysis of these phenomena has revealed the existence of concepts and algorithms encoded within the learned representations of these networks. While significant strides have been made in explaining observed phenomena separately, a unified framework for understanding, dissecting, and predicting the performance of neural networks is lacking. Here, we introduce the semantic landscape paradigm, a conceptual and mathematical framework that describes the training dynamics of neural networks as trajectories on a graph whose nodes correspond to emergent algorithms that are instrinsic to the learned representations of the networks. This abstraction enables us to describe a wide range of neural network phenomena in terms of well studied problems in statistical physics. Specifically, we show that grokking and emergence with scale are associated with percolation phenomena, and neural scaling laws are explainable in terms of the statistics of random walks on graphs. Finally, we discuss how the semantic landscape paradigm complements existing theoretical and practical approaches aimed at understanding and interpreting deep neural networks.

cs.LG↗

arXiv2022

Coupled instabilities drive quasiperiodic order-disorder transitions in Faraday waves

We present an experimental study of quasiperiodic transitions between a highly ordered square-lattice pattern and a disordered, defect-riddled state, in a circular Faraday system. We show that the transition is driven initially by a long-wave amplitude modulation instability, which excites the oscillatory transition phase instability, leading to the formation of dislocations in the Faraday lattice. The appearance of dislocations damps amplitude modulations, which prevents further defects from being created and allows the system to relax back to its ordered state. The process then repeats itself in a quasiperiodic manner. Our experiments reveal a surprising coupling between two distinct instabilities in the Faraday system, and suggest that such coupling may provide a generic mechanism for quasiperiodicity in nonlinear driven dissipative systems.

physics.flu-dyn↗

arXiv2022

Odd dynamics of living chiral crystals

Active crystals are highly ordered structures that emerge from the self-organization of motile objects, and have been widely studied in synthetic and bacterial active matter. Whether collective crystallization phenomena can occur in groups of autonomously developing multicellular organisms is currently unknown. Here, we show that swimming starfish embryos spontaneously assemble into chiral crystals that span thousands of spinning organisms and persist for tens of hours. Combining experiments, theory, and simulations, we demonstrate that the formation, dynamics, and dissolution of these living crystals are controlled by the hydrodynamic properties and natural development of embryos. Remarkably, living chiral crystals exhibit self-sustained chiral oscillations as well as various unconventional deformation response behaviors recently predicted for odd elastic materials. Our results provide direct experimental evidence for how nonreciprocal interactions between autonomous multicellular components may facilitate novel nonequilibrium phases of chiral active matter.

cond-mat.soft↗

arXiv2021

Dynamic clustering of passive colloids in dense suspensions of motile bacteria

Mixtures of active and passive particles are predicted to exhibit a variety of nonequilibrium phases. Here we report a dynamic clustering phase in mixtures of colloids and motile bacteria. We show that colloidal clustering results from a balance between bond breaking due to persistent active motion and bond stabilization due to torques that align active particle velocity tangentially to the passive particle surface. Furthermore, dynamic clustering spans a broad regime between diffusivity-based and motility-induced phase separation that subsumes typical bacterial motility parameters.

cond-mat.soft↗

arXiv2019

Fast and Scalable Group Mutual Exclusion

The group mutual exclusion (GME) problem is a generalization of the classical mutual exclusion problem in which every critical section is associated with a type or session. Critical sections belonging to the same session can execute concurrently, whereas critical sections belonging to different sessions must be executed serially. The well-known read-write mutual exclusion problem is a special case of the group mutual exclusion problem. In this work, we present a novel GME algorithm for an asynchronous shared-memory system that, in addition to satisfying lockout freedom, bounded exit and concurrent entering properties, has O(1) step-complexity when the system contains no conflicting requests as well as O(1) space-complexity per GME object when the system contains sufficient number of GME objects. To the best of our knowledge, no existing GME algorithm has O(1) step-complexity for concurrent entering. Moreover, most existing GME algorithms have Ω(n) space complexity per GME object, where n denotes the number of processes in the system. We also show that our GME algorithm can be easily modified to use bounded space variables.

cs.DC↗

arXiv2016

Deconstructing the glass transition through critical experiments on colloids

The glass transition is the most enduring grand-challenge problem in contemporary condensed matter physics. Here, we review the contribution of colloid experiments to our understanding of this problem. First, we briefly outline the success of colloidal systems in yielding microscopic insights into a wide range of condensed matter phenomena. In the context of the glass transition, we demonstrate their utility in revealing the nature of spatial and temporal dynamical heterogeneity. We then discuss the evidence from colloid experiments in favor of various theories of glass formation that has accumulated over the last two decades. In the next section, we expound on the recent paradigm shift in colloid experiments from an exploratory approach to a critical one aimed at distinguishing between predictions of competing frameworks. We demonstrate how this critical approach is aided by the discovery of novel dynamical crossovers within the range accessible to colloid experiments. We also highlight the impact of alternate routes to glass formation such as random pinning, trajectory space phase transitions and replica coupling on current and future research on the glass transition. We conclude our review by listing some key open challenges in glass physics such as the comparison of growing static lengthscales and the preparation of ultrastable glasses, that can be addressed using colloid experiments.

cond-mat.soft↗

arXiv2016

Influence of an amorphous wall on the distribution of localized excitations in a colloidal glass-forming liquid

Elucidating the nature of the glass transition has been the holy grail of condensed matter physics and statistical mechanics for several decades. A phenomenological aspect that makes glass formation a conceptually formidable problem is that structural and dynamic correlations in glass-forming liquids are too subtle to be captured at the level of conventional two-point functions. As a consequence, a host of theoretical techniques, such as quenched amorphous configurations of particles, have been devised and employed in simulations and colloid experiments to gain insights into the mechanisms responsible for these elusive correlations. Very often, though, the analysis of spatio-temporal correlations is performed in the context of a single theoretical framework, and critical comparisons of microscopic predictions of competing theories are thereby lacking. Here, we address this issue by analysing the distribution of localized excitations, which are building blocks of relaxation as per the Dynamical Facilitation (DF) theory, in the presence of an amorphous wall, a construct motivated by the Random First-Order Transition theory (RFOT). We observe that spatial profiles of the concentration of excitations exhibit complex features such as non-monotonicity and oscillations. Moreover, the smoothly varying part of the concentration profile yields a length scale $ξ_c$, which we compare with a previously computed length scale $ξ_{dyn}$. Our results suggest a method to assess the role of dynamical facilitation in governing structural relaxation in glass-forming liquids.

cond-mat.soft↗

arXiv2016

Robust Neighbor Discovery in Multi-Hop Multi-Channel Heterogeneous Wireless Networks

An important first step when deploying a wireless ad hoc network is neighbor discovery in which every node attempts to determine the set of nodes it can communicate with in one wireless hop. In the recent years, cognitive radio (CR) technology has gained attention as an attractive approach to alleviate spectrum congestion. A cognitive radio transceiver can operate over a wide range of frequencies, possibly scanning multiple frequency bands. A cognitive radio node can opportunistically utilize unused wireless spectrum without interference from other wireless devices in its vicinity. Due to spatial variations in frequency usage and hardware variations in radio transceivers, different nodes in the network may perceive different subsets of frequencies available to them for communication. This heterogeneity in the available channel sets across the network increases the complexity of solving the neighbor discovery problem in a cognitive radio network. In this work, we design and analyze several randomized algorithms for neighbor discovery in such a (heterogeneous) network under a variety of assumptions (e.g. maximum node degree known or unknown) for both synchronous and asynchronous systems under minimal knowledge. We also show that our randomized algorithms are naturally suited to tolerate unreliable channels and adversarial attacks.

cs.DC↗

arXiv2014

Direct measurements of growing amorphous order and non-monotonic dynamic correlations in a colloidal glass-former

While the transformation of flowing liquids into rigid glasses is omnipresent, a complete understanding of vitrification remains elusive. Of the numerous approaches aimed at solving the glass transition problem, the Random First-Order Theory (RFOT) is the most prominent. However, the existence of the underlying thermodynamic phase transition envisioned by RFOT remains debatable, since its key microscopic predictions concerning the growth of amorphous order and the nature of dynamic correlations lack experimental verification. Here, by using holographic optical tweezers, we freeze a wall of particles in an equilibrium configuration of a 2D colloidal glass-forming liquid and provide direct evidence for growing amorphous order in the form of a static point-to-set length. Most remarkably, we uncover the non-monotonic dependence of dynamic correlations on area fraction and show that this non-monotonicity follows directly from the change in morphology of cooperatively rearranging regions, as predicted by RFOT. Our findings suggest that the glass transition has a thermodynamic origin.

cond-mat.soft↗

arXiv2014

Dynamical facilitation governs glassy dynamics in suspensions of colloidal ellipsoids

One of the greatest challenges in contemporary condensed matter physics is to ascertain whether the formation of glasses from liquids is fundamentally thermodynamic or dynamic in origin. While the thermodynamic paradigm has dominated theoretical research for decades, the purely kinetic perspective of the dynamical facilitation (DF) theory has attained prominence in recent times. In particular, recent experiments and simulations have highlighted the importance of facilitation using simple model systems composed of spherical particles. However, an overwhelming majority of liquids possess anisotropy in particle shape and interactions and it is therefore imperative to examine facilitation in complex glass-formers. Here, we apply the DF theory to systems with orientational degrees of freedom as well as anisotropic attractive interactions. By analyzing data from experiments on colloidal ellipsoids, we show that facilitation plays a pivotal role in translational as well as orientational relaxation. Further, we demonstrate that the introduction of attractive interactions leads to spatial decoupling of translational and rotational facilitation, which subsequently results in the decoupling of dynamical heterogeneities. Most strikingly, the DF theory can predict the existence of reentrant glass transitions based on the statistics of localized dynamical events, called excitations, whose duration is substantially smaller than the structural relaxation time. Our findings pave the way for systematically testing the DF approach in complex glass-formers and also establish the significance of facilitation in governing structural relaxation in supercooled liquids.

cond-mat.soft↗

arXiv2014

Growing Dynamical Facilitation on Approaching the Random Pinning Colloidal Glass Transition

Despite decades of research, it remains to be established whether the transformation of a liquid into a glass is fundamentally thermodynamic or dynamic in origin. While observations of growing length scales are consistent with thermodynamic perspectives like the Random First-Order Transition theory (RFOT), the purely dynamic approach of the Dynamical Facilitation (DF) theory lacks experimental validation. Further, for glass transitions induced by randomly freezing a subset of particles in the liquid phase, simulations support the predictions of RFOT, whereas the DF theory remains unexplored. Here, using video microscopy and holographic optical tweezers, we show that dynamical facilitation in a colloidal glass-forming liquid unambiguously grows with density as well as the fraction of pinned particles. In addition, we show that heterogeneous dynamics in the form of string-like cooperative motion, which is believed to be consistent with RFOT, emerges naturally within the framework of facilitation. Most importantly, our findings demonstrate that a purely dynamic origin of the glass transition cannot be ruled out.

cond-mat.soft↗

arXiv2014

Direct Evidence for an Absorbing Phase Transition Governing Yielding of a Soft Glass

We present the first direct experimental evidence showing that yielding of a prototypical soft solid - a colloidal glass - is a non-equilibrium 'absorbing' phase transition. By simultaneously quantifying single-particle dynamics and bulk mechanical response, we extracted critical exponents for the order parameter and the relaxation time and found that this transition belongs to the conserved directed percolation universality class. In addition, the observed critical slowing down is accompanied by a growing correlation length associated with the size of regions of high Debye-Waller factor which are precursors to yield events in glasses. Our results show unambiguously that yielding of soft solids falls squarely in the realm of non-equilibrium critical phenomena.

cond-mat.soft↗