arXiv subjects

Shaan Desai

Publications and source records attributed to Shaan Desai.

10 recordsLinked to original sources

arXiv2026

Think in English, Answer in Korean: Efficient Adaptation of Multilingual Tool-Using Agents

We present LuckyStar 111B, a 111B-parameter hybrid reasoning model developed through a collaboration between Cohere and LG CNS for Korean-English enterprise agents under practical memory and serving constraints. The model trains from Cohere's fully post-trained Command A model rather than a new pretraining run, and uses preamble conditioning to switch between concise non-reasoning behavior and longer tool-oriented reasoning. We study four choices for scaling tool-using agents efficiently: multilingual supervised fine-tuning, reinforcement learning with verifiable rewards for multi-step tool-use tasks, language-consistency rewards for Korean user-facing responses, and 4-bit quantization for single-GPU serving. The adapted model improves mathematical reasoning, function calling, and agentic natural-language-to-SQL (NL2SQL) performance while preserving general Korean and English instruction-following quality. These results provide a practical recipe and failure-mode analysis for adapting post-trained multilingual models to verifiable agentic workflows under memory-constrained deployment.

cs.AI↗

arXiv2025

Command A: An Enterprise-Ready Large Language Model

In this report we describe the development of Command A, a powerful large language model purpose-built to excel at real-world enterprise use cases. Command A is an agent-optimised and multilingual-capable model, with support for 23 languages of global business, and a novel hybrid architecture balancing efficiency with top of the range performance. It offers best-in-class Retrieval Augmented Generation (RAG) capabilities with grounding and tool use to automate sophisticated business processes. These abilities are achieved through a decentralised training approach, including self-refinement algorithms and model merging techniques. We also include results for Command R7B which shares capability and architectural similarities to Command A. Weights for both models have been released for research purposes. This technical report details our original training pipeline and presents an extensive evaluation of our models across a suite of enterprise-relevant tasks and public benchmarks, demonstrating excellent performance and efficiency.

cs.CL↗

arXiv2025

Command R7B Arabic: A Small, Enterprise Focused, Multilingual, and Culturally Aware Arabic LLM

Building high-quality large language models (LLMs) for enterprise Arabic applications remains challenging due to the limited availability of digitized Arabic data. In this work, we present a data synthesis and refinement strategy to help address this problem, namely, by leveraging synthetic data generation and human-in-the-loop annotation to expand our Arabic training corpus. We further present our iterative post training recipe that is essential to achieving state-of-the-art performance in aligning the model with human preferences, a critical aspect to enterprise use cases. The culmination of this effort is the release of a small, 7B, open-weight model that outperforms similarly sized peers in head-to-head comparisons and on Arabic-focused benchmarks covering cultural knowledge, instruction following, RAG, and contextual faithfulness.

cs.CL↗

arXiv2022

One-Shot Transfer Learning of Physics-Informed Neural Networks

Solving differential equations efficiently and accurately sits at the heart of progress in many areas of scientific research, from classical dynamical systems to quantum mechanics. There is a surge of interest in using Physics-Informed Neural Networks (PINNs) to tackle such problems as they provide numerous benefits over traditional numerical approaches. Despite their potential benefits for solving differential equations, transfer learning has been under explored. In this study, we present a general framework for transfer learning PINNs that results in one-shot inference for linear systems of both ordinary and partial differential equations. This means that highly accurate solutions to many unknown differential equations can be obtained instantaneously without retraining an entire network. We demonstrate the efficacy of the proposed deep learning approach by solving several real-world problems, such as first- and second-order linear ordinary equations, the Poisson equation, and the time-dependent Schrodinger complex-value partial differential equation.

cs.LG↗

arXiv2021

Port-Hamiltonian Neural Networks for Learning Explicit Time-Dependent Dynamical Systems

Accurately learning the temporal behavior of dynamical systems requires models with well-chosen learning biases. Recent innovations embed the Hamiltonian and Lagrangian formalisms into neural networks and demonstrate a significant improvement over other approaches in predicting trajectories of physical systems. These methods generally tackle autonomous systems that depend implicitly on time or systems for which a control signal is known apriori. Despite this success, many real world dynamical systems are non-autonomous, driven by time-dependent forces and experience energy dissipation. In this study, we address the challenge of learning from such non-autonomous systems by embedding the port-Hamiltonian formalism into neural networks, a versatile framework that can capture energy dissipation and time-dependent control forces. We show that the proposed \emph{port-Hamiltonian neural network} can efficiently learn the dynamics of nonlinear physical systems of practical interest and accurately recover the underlying stationary Hamiltonian, time-dependent force, and dissipative coefficient. A promising outcome of our network is its ability to learn and predict chaotic systems such as the Duffing equation, for which the trajectories are typically hard to learn.

cs.LG↗

arXiv2021

Variational Integrator Graph Networks for Learning Energy Conserving Dynamical Systems

Recent advances show that neural networks embedded with physics-informed priors significantly outperform vanilla neural networks in learning and predicting the long term dynamics of complex physical systems from noisy data. Despite this success, there has only been a limited study on how to optimally combine physics priors to improve predictive performance. To tackle this problem we unpack and generalize recent innovations into individual inductive bias segments. As such, we are able to systematically investigate all possible combinations of inductive biases of which existing methods are a natural subset. Using this framework we introduce Variational Integrator Graph Networks - a novel method that unifies the strengths of existing approaches by combining an energy constraint, high-order symplectic variational integrators, and graph neural networks. We demonstrate, across an extensive ablation, that the proposed unifying framework outperforms existing methods, for data-efficient learning and in predictive accuracy, across both single and many-body problems studied in recent literature. We empirically show that the improvements arise because high order variational integrators combined with a potential energy constraint induce coupled learning of generalized position and momentum updates which can be formalized via the Partitioned Runge-Kutta method.

cs.LG↗

arXiv2021

Tuning Mixed Input Hyperparameters on the Fly for Efficient Population Based AutoRL

Despite a series of recent successes in reinforcement learning (RL), many RL algorithms remain sensitive to hyperparameters. As such, there has recently been interest in the field of AutoRL, which seeks to automate design decisions to create more general algorithms. Recent work suggests that population based approaches may be effective AutoRL algorithms, by learning hyperparameter schedules on the fly. In particular, the PB2 algorithm is able to achieve strong performance in RL tasks by formulating online hyperparameter optimization as time varying GP-bandit problem, while also providing theoretical guarantees. However, PB2 is only designed to work for continuous hyperparameters, which severely limits its utility in practice. In this paper we introduce a new (provably) efficient hierarchical approach for optimizing both continuous and categorical variables, using a new time-varying bandit algorithm specifically designed for the population based training regime. We evaluate our approach on the challenging Procgen benchmark, where we show that explicitly modelling dependence between data augmentation and other hyperparameters improves generalization.

cs.LG↗

arXiv2020

DYNOTEARS: Structure Learning from Time-Series Data

We revisit the structure learning problem for dynamic Bayesian networks and propose a method that simultaneously estimates contemporaneous (intra-slice) and time-lagged (inter-slice) relationships between variables in a time-series. Our approach is score-based, and revolves around minimizing a penalized loss subject to an acyclicity constraint. To solve this problem, we leverage a recent algebraic result characterizing the acyclicity constraint as a smooth equality constraint. The resulting algorithm, which we call DYNOTEARS, outperforms other methods on simulated data, especially in high-dimensions as the number of variables increases. We also apply this algorithm on real datasets from two different domains, finance and molecular biology, and analyze the resulting output. Compared to state-of-the-art methods for learning dynamic Bayesian networks, our method is both scalable and accurate on real data. The simple formulation and competitive performance of our method make it suitable for a variety of problems where one seeks to learn connections between variables across time.

stat.ML↗

arXiv2019

Deep learning-based quality filtering of mechanically exfoliated 2D crystals

Two-dimensional (2D) crystals are attracting growing interest in various research fields such as engineering, physics, chemistry, pharmacy and biology owing to their low dimensionality and dramatic change of properties compared to the bulk counterparts. Among the various techniques used to manufacture 2D crystals, mechanical exfoliation has been essential to practical applications and fundamental research. However, mechanically exfoliated crystals on substrates contain relatively thick flakes that must be found and removed manually, limiting high-throughput manufacturing of atomic 2D crystals and van der Waals heterostructures. Here we present a deep learning-based method to segment and identify the thickness of atomic layer flakes from optical microscopy images. Through carefully designing a neural network based on U-Net, we found that our neural network based on U-net trained only with the data based on 24 images successfully distinguish monolayer and bilayer MoS2 with a success rate of 70%, which is a practical value in the first screening process for choosing monolayer and bilayer flakes of MoS2 of all flakes on substrates without human eye. The remarkable results highlight the possibility that a large fraction of manual laboratory work can be replaced by AI-based systems, boosting productivity.

physics.app-ph↗

arXiv2018

Data-driven studies of magnetic two-dimensional materials

We use a data-driven approach to study the magnetic and thermodynamic properties of van der Waals (vdW) layered materials. We investigate monolayers of the form A$_2$B$_2$X$_6$, based on the known material Cr$_2$Ge$_2$Te$_6$, using density functional theory (DFT) calculations and machine learning methods to determine their magnetic properties, such as magnetic order and magnetic moment. We also examine formation energies and use them as a proxy for chemical stability. We show that machine learning tools, combined with DFT calculations, can provide a computationally efficient means to predict properties of such two-dimensional (2D) magnetic materials. Our data analytics approach provides insights into the microscopic origins of magnetic ordering in these systems. For instance, we find that the X site strongly affects the magnetic coupling between neighboring A sites, which drives the magnetic ordering. Our approach opens new ways for rapid discovery of chemically stable vdW materials that exhibit magnetic behavior.

cond-mat.mtrl-sci↗