Searcharxiv⌕ Search

arXiv subjects

Saurabh Kumar

Publications and source records attributed to Saurabh Kumar.

At least 37 records · Page 2Linked to original sources

Influence of Hydrogen on Dislocation Relaxation in BCC Iron: Atomistic Mechanisms and Implications

In this study, the influence of pure dislocation and hydrogen-dislocation interactions on anelastic response or internal friction relaxation peaks in bcc-iron was investigated. These relaxations are primarily governed by thermally activated kink nucleation and kink migration events. An atomistic multiscale framework, coupling molecular dynamics (MD) and kinetic Monte Carlo (KMC) simulations, was developed to investigate the underlying atomistic mechanisms behind dislocation-relaxation peaks. MD simulations revealed that the presence of hydrogen atoms near the dislocation core facilitates the kink nucleation process by reducing the nucleation barrier while enhancing the barrier for dislocation migration. The KMC model captured Snoek-Koster peaks arising from the Cottrell atmosphere formed by hydrogen atoms and clusters around the dislocation core, providing insights into the atomistic mechanisms controlling these relaxations. Furthermore, the proposed computational scheme elucidated a unique linear relationship between hydrogen content and the internal friction loss factor, offering a methodology for hydrogen detection and quantification.

cond-mat.mtrl-sci↗

State-of-the-art Small Language Coder Model: Mify-Coder

We present Mify-Coder, a 2.5B-parameter code model trained on 4.2T tokens using a compute-optimal strategy built on the Mify-2.5B foundation model. Mify-Coder achieves comparable accuracy and safety while significantly outperforming much larger baseline models on standard coding and function-calling benchmarks, demonstrating that compact models can match frontier-grade models in code generation and agent-driven workflows. Our training pipeline combines high-quality curated sources with synthetic data generated through agentically designed prompts, refined iteratively using enterprise-grade evaluation datasets. LLM-based quality filtering further enhances data density, enabling frugal yet effective training. Through disciplined exploration of CPT-SFT objectives, data mixtures, and sampling dynamics, we deliver frontier-grade code intelligence within a single continuous training trajectory. Empirical evidence shows that principled data and compute discipline allow smaller models to achieve competitive accuracy, efficiency, and safety compliance. Quantized variants of Mify-Coder enable deployment on standard desktop environments without requiring specialized hardware.

cs.SE↗

Cooperative Nonlinear Guidance Strategies for Guaranteed Pursuit-Evasion

This paper investigates a pursuit-evasion problem involving three agents: a pursuer, an evader, and a defender. Cooperative guidance laws are developed for the evader-defender team that guarantee interception of the pursuer by the defender before it reaches the vicinity of the evader. Unlike heuristic methods, optimal control, differential game formulation, and recently proposed time-constrained guidance techniques, a geometry-based solution is proposed to safeguard the evader from the pursuer's incoming threat. The proposed strategy is computationally efficient and expected to be scalable as the number of agents increases. Another notable feature of the proposed strategy is that the evader-defender team does not require knowledge of the pursuer's strategy, yet the pursuer's interception is guaranteed for arbitrary initial engagement geometries. It is further shown that the relevant error variables for the evader-defender team (or individual) converge to zero at a prespecified finite time that can be exactly prescribed prior to the three-body engagement. Finally, the effectiveness of the proposed cooperative pursuit-evasion strategy is demonstrated through simulations across diverse engagement scenarios.

eess.SY↗

Shadows of generalised Hayward spacetimes : in vacuum and with plasma

We investigate the shadow properties of a wide class of spacetimes arising from different parameter regimes of the generalized Hayward metric, characterized by two independent parameters $(σ, κ)$ (Phys. Rev. D 106, 044028). This metric extends the original Hayward regular black hole solution by introducing distinct mass functions in the $g_{tt}$ and $g_{rr}$ components, giving rise to four types of wormholes ( which include multi-peak effective potentials), a regular black hole, and a singular black hole solutions allowing for a unified treatment of black hole mimickers. We compute the shadow radii for all spacetimes in vacuum and in the presence of plasma, using both homogeneous and non-homogeneous plasma profiles. Our results show that certain wormhole solutions particularly the Hayward-Damour-Solodukhin class can exhibit multiple photon spheres, leading to shadow features that differ significantly from the Schwarzschild black hole. When these results are compared with Event Horizon Telescope observations of Sgr~$A^\star$, we find that regular black holes remain observationally viable but only within a narrow parameter space. In contrast, wormhole solutions with multi-peak effective potentials are more consistent with shadow constraints than those with single peaks. This contrasts with quasinormal mode studies, which favored single-barrier potentials, and may imply detectable late-time echoes in gravitational wave signals.

gr-qc↗

Continual Learning as Computationally Constrained Reinforcement Learning

An agent that efficiently accumulates knowledge to develop increasingly sophisticated skills over a long lifetime could advance the frontier of artificial intelligence capabilities. The design of such agents, which remains a long-standing challenge of artificial intelligence, is addressed by the subject of continual learning. This monograph clarifies and formalizes concepts of continual learning, introducing a framework and set of tools to stimulate further research.

cs.LG↗

Hierarchical Text Classification Using Contrastive Learning Informed Path Guided Hierarchy

Hierarchical Text Classification (HTC) has recently gained traction given the ability to handle complex label hierarchy. This has found applications in domains like E- commerce, customer care and medicine industry among other real-world applications. Existing HTC models either encode label hierarchy separately and mix it with text encoding or guide the label hierarchy structure in the text encoder. Both approaches capture different characteristics of label hierarchy and are complementary to each other. In this paper, we propose a Hierarchical Text Classification using Contrastive Learning Informed Path guided hierarchy (HTC-CLIP), which learns hierarchy-aware text representation and text informed path guided hierarchy representation using contrastive learning. During the training of HTC-CLIP, we learn two different sets of class probabilities distributions and during inference, we use the pooled output of both probabilities for each class to get the best of both representations. Our results show that the two previous approaches can be effectively combined into one architecture to achieve improved performance. Tests on two public benchmark datasets showed an improvement of 0.99 - 2.37% in Macro F1 score using HTC-CLIP over the existing state-of-the-art models.

cs.CL↗

Building a Few-Shot Cross-Domain Multilingual NLU Model for Customer Care

Customer care is an essential pillar of the e-commerce shopping experience with companies spending millions of dollars each year, employing automation and human agents, across geographies (like US, Canada, Mexico, Chile), channels (like Chat, Interactive Voice Response (IVR)), and languages (like English, Spanish). SOTA pre-trained models like multilingual-BERT, fine-tuned on annotated data have shown good performance in downstream tasks relevant to Customer Care. However, model performance is largely subject to the availability of sufficient annotated domain-specific data. Cross-domain availability of data remains a bottleneck, thus building an intent classifier that generalizes across domains (defined by channel, geography, and language) with only a few annotations, is of great practical value. In this paper, we propose an embedder-cum-classifier model architecture which extends state-of-the-art domain-specific models to other domains with only a few labeled samples. We adopt a supervised fine-tuning approach with isotropic regularizers to train a domain-specific sentence embedder and a multilingual knowledge distillation strategy to generalize this embedder across multiple domains. The trained embedder, further augmented with a simple linear classifier can be deployed for new domains. Experiments on Canada and Mexico e-commerce Customer Care dataset with few-shot intent detection show an increase in accuracy by 20-23% against the existing state-of-the-art pre-trained models.

cs.CL↗

Conformal Transformations for Symmetric Power Transformers

Transformers with linear attention offer significant computational advantages over softmax-based transformers but often suffer from degraded performance. The symmetric power (sympow) transformer, a particular type of linear transformer, addresses some of this performance gap by leveraging symmetric tensor embeddings, achieving comparable performance to softmax transformers. However, the finite capacity of the recurrent state in sympow transformers limits their ability to retain information, leading to performance degradation when scaling the training or evaluation context length. To address this issue, we propose the conformal-sympow transformer, which dynamically frees up capacity using data-dependent multiplicative gating and adaptively stores information using data-dependent rotary embeddings. Preliminary experiments on the LongCrawl64 dataset demonstrate that conformal-sympow overcomes the limitations of sympow transformers, achieving robust performance across scaled training and evaluation contexts.

cs.LG↗

Provably Safe Control for Constrained Nonlinear Systems with Bounded Input

In real-world control applications, actuator constraints and output constraints (specifically in tracking problems) are inherent and critical to ensuring safe and reliable operation. However, generally, control strategies often neglect these physical limitations, leading to potential instability, degraded performance, or even system failure when deployed on real-world systems. This paper addresses the control design problem for a class of nonlinear systems under both actuator saturation and output constraints. First, a smooth asymmetric saturation model (a more generic representative of practical scenarios) is proposed to model actuator saturation, which ensures that the control inputs always remain confined within a predefined set to ensure safety. Based on the proposed model, we develop a nonlinear control framework that guarantees output tracking while ensuring that system output remains confined to the predefined set. Later, we integrate this design with the constrained output tracking control problem, wherein we show that the system output tracks its desired trajectory by simultaneously satisfying input and output constraints. The global stabilization of the tracking error is achieved in the presence of input constraints, while semi-global stabilization is achieved in the presence of both input and output constraints. Additionally, we rigorously establish the boundedness of all closed-loop signals under the proposed design. Simulation results demonstrate the effectiveness of the proposed methods in handling asymmetric constraints while achieving desirable tracking performance.

eess.SY↗

Three-dimensional Nonlinear Path-following Guidance with Bounded Input Constraints

In this paper, we consider the tracking of arbitrary curvilinear geometric paths in three-dimensional output spaces of unmanned aerial vehicles (UAVs) without pre-specified timing requirements, commonly referred to as path-following problems, subjected to bounded inputs. Specifically, we propose a novel nonlinear path-following guidance law for a UAV that enables it to follow any smooth curvilinear path in three dimensions while accounting for the bounded control authority in the design. The proposed solution offers a general treatment of the path-following problem by removing the dependency on the path's geometry, which makes it applicable to paths with varying levels of complexity and smooth curvatures. Additionally, the proposed strategy draws inspiration from the pursuit guidance approach, which is known for its simplicity and ease of implementation. Theoretical analysis guarantees that the UAV converges to its desired path within a fixed time and remains on it irrespective of its initial configuration with respect to the path. Finally, the simulations demonstrate the merits and effectiveness of the proposed guidance strategy through a wide range of engagement scenarios, showcasing the UAV's ability to follow diverse curvilinear paths accurately.

eess.SY↗

SCRUBD: Smart Contracts Reentrancy and Unhandled Exceptions Vulnerability Dataset

Smart Contracts (SCs) handle transactions in the Ethereum blockchain worth millions of United States dollars, making them a lucrative target for attackers seeking to exploit vulnerabilities and steal funds. The Ethereum community has developed a rich set of tools to detect vulnerabilities in SCs, including reentrancy (RE) and unhandled exceptions (UX). A dataset of SCs labelled with vulnerabilities is needed to evaluate the tools' efficacy. Existing SC datasets with labelled vulnerabilities have limitations, such as covering only a limited range of vulnerability scenarios and containing incorrect labels. As a result, there is a lack of a standardized dataset to compare the performances of these tools. SCRUBD aims to fill this gap. We present a dataset of real-world SCs and synthesized SCs labelled with RE and UX. The real-world SC dataset is labelled through crowdsourcing, followed by manual inspection by an expert, and covers both RE and UX vulnerabilities. On the other hand, the synthesized dataset is carefully crafted to cover various RE scenarios only. Using SCRUBD we compared the performance of six popular vulnerability detection tools. Based on our study, we found that Slither outperforms other tools on a crowdsourced dataset in detecting RE vulnerabilities, while Sailfish outperforms other tools on a manually synthesized dataset for detecting RE. For UX vulnerabilities, Slither outperforms all other tools.

cs.CR↗

Learning Continually by Spectral Regularization

Loss of plasticity is a phenomenon where neural networks can become more difficult to train over the course of learning. Continual learning algorithms seek to mitigate this effect by sustaining good performance while maintaining network trainability. We develop a new technique for improving continual learning inspired by the observation that the singular values of the neural network parameters at initialization are an important factor for trainability during early phases of learning. From this perspective, we derive a new spectral regularizer for continual learning that better sustains these beneficial initialization properties throughout training. In particular, the regularizer keeps the maximum singular value of each layer close to one. Spectral regularization directly ensures that gradient diversity is maintained throughout training, which promotes continual trainability, while minimally interfering with performance in a single task. We present an experimental analysis that shows how the proposed spectral regularizer can sustain trainability and performance across a range of model architectures in continual supervised and reinforcement learning settings. Spectral regularization is less sensitive to hyperparameters while demonstrating better training in individual tasks, sustaining trainability as new tasks arrive, and achieving better generalization performance.

cs.LG↗

Maintaining Plasticity in Continual Learning via Regenerative Regularization

In continual learning, plasticity refers to the ability of an agent to quickly adapt to new information. Neural networks are known to lose plasticity when processing non-stationary data streams. In this paper, we propose L2 Init, a simple approach for maintaining plasticity by incorporating in the loss function L2 regularization toward initial parameters. This is very similar to standard L2 regularization (L2), the only difference being that L2 regularizes toward the origin. L2 Init is simple to implement and requires selecting only a single hyper-parameter. The motivation for this method is the same as that of methods that reset neurons or parameter values. Intuitively, when recent losses are insensitive to particular parameters, these parameters should drift toward their initial values. This prepares parameters to adapt quickly to new tasks. On problems representative of different types of nonstationarity in continual supervised learning, we demonstrate that L2 Init most consistently mitigates plasticity loss compared to previously proposed approaches.

cs.LG↗

The Need for a Big World Simulator: A Scientific Challenge for Continual Learning

The "small agent, big world" frame offers a conceptual view that motivates the need for continual learning. The idea is that a small agent operating in a much bigger world cannot store all information that the world has to offer. To perform well, the agent must be carefully designed to ingest, retain, and eject the right information. To enable the development of performant continual learning agents, a number of synthetic environments have been proposed. However, these benchmarks suffer from limitations, including unnatural distribution shifts and a lack of fidelity to the "small agent, big world" framing. This paper aims to formalize two desiderata for the design of future simulated environments. These two criteria aim to reflect the objectives and complexity of continual learning in practical settings while enabling rapid prototyping of algorithms on a smaller scale.

cs.LG↗

Satisficing Exploration for Deep Reinforcement Learning

A default assumption in the design of reinforcement-learning algorithms is that a decision-making agent always explores to learn optimal behavior. In sufficiently complex environments that approach the vastness and scale of the real world, however, attaining optimal performance may in fact be an entirely intractable endeavor and an agent may seldom find itself in a position to complete the requisite exploration for identifying an optimal policy. Recent work has leveraged tools from information theory to design agents that deliberately forgo optimal solutions in favor of sufficiently-satisfying or satisficing solutions, obtained through lossy compression. Notably, such agents may employ fundamentally different exploratory decisions to learn satisficing behaviors more efficiently than optimal ones that are more data intensive. While supported by a rigorous corroborating theory, the underlying algorithm relies on model-based planning, drastically limiting the compatibility of these ideas with function approximation and high-dimensional observations. In this work, we remedy this issue by extending an agent that directly represents uncertainty over the optimal value function allowing it to both bypass the need for model-based planning and to learn satisficing policies. We provide simple yet illustrative experiments that demonstrate how our algorithm enables deep reinforcement-learning agents to achieve satisficing behaviors. In keeping with previous work on this setting for multi-armed bandits, we additionally find that our algorithm is capable of synthesizing optimal behaviors, when feasible, more efficiently than its non-information-theoretic counterpart.

cs.LG↗

Trajectory Tracking for Unmanned Aerial Vehicles in 3D Spaces under Motion Constraints

This article presents a three-dimensional nonlinear trajectory tracking control strategy for unmanned aerial vehicles (UAVs) in the presence of spatial constraints. As opposed to many existing control strategies, which do not consider spatial constraints, the proposed strategy considers spatial constraints on each degree of freedom movement of the UAV. Such consideration makes the design appealing for many practical applications, such as pipeline inspection, boundary tracking, etc. The proposed design accounts for the limited information about the inertia matrix, thereby affirming its inherent robustness against unmodeled dynamics and other imperfections. We rigorously show that the UAV will converge to its desired path by maintaining bounded position, orientation, and linear and angular speeds. Finally, we demonstrate the effectiveness of the proposed strategy through various numerical simulations.

eess.SY↗

Observational signatures of Rotating compact objects in Plasma space-time

We have investigated the characteristics of shadows cast by the Kerr black hole in the presence of plasma and compared them to those of a rotating wormhole in a uniform plasma space-time for an observer at infinity. Interestingly, for the same uniform plasma density, the apparent shadow size of the rotating wormhole is always greater than that of the Kerr black hole. To further distinguish the two compact objects we studied the deflection angle and did a comparative study in the presence of the uniform and non-uniform plasma profiles. The goal of this whole exercise is to deepen our understanding of the observational phenomena of these astrophysical objects. The analysis reveals the importance of specific plasma distribution profiles, the impact of plasma on the shadow diameter, and the behavior of deflection angles in different plasma scenarios. We have calculated constraints on the plasma parameters by considering observational data and employing analytical formulations. Our work therefore provides valuable insights into the behavior of light rays near compact objects in plasma space-time.

gr-qc↗

Speaking rate attention-based duration prediction for speed control TTS

With the advent of high-quality speech synthesis, there is a lot of interest in controlling various prosodic attributes of speech. Speaking rate is an essential attribute towards modelling the expressivity of speech. In this work, we propose a novel approach to control the speaking rate for non-autoregressive TTS. We achieve this by conditioning the speaking rate inside the duration predictor, allowing implicit speaking rate control. We show the benefits of this approach by synthesising audio at various speaking rate factors and measuring the quality of speaking rate-controlled synthesised speech. Further, we study the effect of the speaking rate distribution of the training data towards effective rate control. Finally, we fine-tune a baseline pretrained TTS model to obtain speaking rate control TTS. We provide various analyses to showcase the benefits of using this proposed approach, along with objective as well as subjective metrics. We find that the proposed methods have higher subjective scores and lower speaker rate errors across many speaking rate factors over the baseline.

eess.AS↗