arXiv subjects

Amandeep Singh

Publications and source records attributed to Amandeep Singh.

At least 19 recordsLinked to original sources

arXiv2026

HumaniBench: A Human-Centric Framework for Large Multimodal Models Evaluation

Although recent large multimodal models (LMMs) show impressive progress on vision language tasks, their alignment with human centered (HC) principles such as fairness, ethics, inclusivity, empathy, and robustness is often overlooked. Existing LMM benchmarks are largely accuracy-agnostic. We present HumaniBench, a unified framework for characterizing HC alignment across realistic, socially grounded visual contexts. It contains 32,000 expert-verified image-question pairs from real-world news imagery, each mapped to one or more HC principles through explicit metrics. Comparing 15 state of the art LMMs reveals consistent trade -offs: proprietary systems lead on ethics, reasoning, and empathy, while open-source models show superior visual grounding and resilience. All models show persistent gaps in fairness and multilingual inclusivity. Chain-of-thought prompting and test-time scaling yield 8to 12 % gains on several HC dimensions. HumaniBench enables fine-grained analysis of alignment trade-offs not captured by conventional multimodal benchmarks. https://vectorinstitute.github.io/humanibench/

cs.CV↗

arXiv2026

Magnetic Microscopy of Skyrmions in Magnetic Thin Films with Chiral Overlayers

Topologically nontrivial magnetic textures such as skyrmions offer promising opportunities for spintronic applications. In recent years, it has been shown that the magnetic properties of layered materials can be affected by depositing chiral molecules on the surface, while the influence of chiral overlayers on skyrmion properties such as their stability and interactions remains largely unexplored. To address this challenge, we employ wide-field nitrogen-vacancy (NV) magnetometry to directly image skyrmions in chiral-molecule-functionalized magnetic thin films, enabling quantitative mapping of magnetic stray fields over extended areas under ambient conditions. Using pixel-resolved optically detected magnetic resonance (ODMR) combined with controlled magnetic fields, we reproducibly nucleate and probe skyrmion states in CoFeB ferromagnetic samples, enabling quantitative investigation of their properties. We find evidence for enantioselective and magnetic-field-polarity-dependent modifications of skyrmion diameter, spacing, and shape, pointing to a possibility of molecular control of topological spin textures via magneto-chiral coupling.

cond-mat.mtrl-sci↗

arXiv2025

Quantum Imaging of Ferromagnetic van der Waals Magnetic Domain Structures at Ambient Conditions

Recently discovered 2D van der Waals magnetic materials, and specifically Iron-Germanium-Telluride ($\rm Fe_{5}GeTe_{2}$), have attracted significant attention both from a fundamental perspective and for potential applications. Key open questions concern their domain structure and magnetic phase transition temperature as a function of sample thickness and external field, as well as implications for integration into devices such as magnetic memories and logic. Here we address key questions using a nitrogen-vacancy center based quantum magnetic microscope, enabling direct imaging of the magnetization of $\rm Fe_{5}GeTe_{2}$ at sub-micron spatial resolution as a function of temperature, magnetic field, and thickness. We employ spatially resolved measures, including magnetization variance and cross-correlation, and find a significant spread in transition temperature yet with no clear dependence on thickness down to 15 nm. We also identify previously unknown stripe features in the optical as well as magnetic images, which we attribute to modulations of the constituting elements during crystal synthesis and subsequent oxidation. Our results suggest that the magnetic anisotropy in this material does not play a crucial role in their magnetic properties, leading to a magnetic phase transition of $\rm Fe_{5}GeTe_{2}$ which is largely thickness-independent down to 15 nm. Our findings could be significant in designing future spintronic devices, magnetic memories and logic with 2D van der Waals magnetic materials.

cond-mat.mtrl-sci↗

arXiv2025

FedRAG: A Framework for Fine-Tuning Retrieval-Augmented Generation Systems

Retrieval-augmented generation (RAG) systems have been shown to be effective in addressing many of the drawbacks of relying solely on the parametric memory of large language models. Recent work has demonstrated that RAG systems can be improved via fine-tuning of their retriever and generator models. In this work, we introduce FedRAG, a framework for fine-tuning RAG systems across centralized and federated architectures. FedRAG supports state-of-the-art fine-tuning methods, offering a simple and intuitive interface and a seamless conversion from centralized to federated training tasks. FedRAG is also deeply integrated with the modern RAG ecosystem, filling a critical gap in available tools.

cs.LG↗

arXiv2025

Spatiotemporal deep learning models for detection of rapid intensification in cyclones

Cyclone rapid intensification is the rapid increase in cyclone wind intensity, exceeding a threshold of 30 knots, within 24 hours. Rapid intensification is considered an extreme event during a cyclone, and its occurrence is relatively rare, contributing to a class imbalance in the dataset. A diverse array of factors influences the likelihood of a cyclone undergoing rapid intensification, further complicating the task for conventional machine learning models. In this paper, we evaluate deep learning, ensemble learning and data augmentation frameworks to detect cyclone rapid intensification based on wind intensity and spatial coordinates. We note that conventional data augmentation methods cannot be utilised for generating spatiotemporal patterns replicating cyclones that undergo rapid intensification. Therefore, our framework employs deep learning models to generate spatial coordinates and wind intensity that replicate cyclones to address the class imbalance problem of rapid intensification. We also use a deep learning model for the classification module within the data augmentation framework to differentiate between rapid and non-rapid intensification events during a cyclone. Our results show that data augmentation improves the results for rapid intensification detection in cyclones, and spatial coordinates play a critical role as input features to the given models. This paves the way for research in synthetic data generation for spatiotemporal data with extreme events.

cs.LG↗

arXiv2024

Hybrid Forecasting of Geopolitical Events

Sound decision-making relies on accurate prediction for tangible outcomes ranging from military conflict to disease outbreaks. To improve crowdsourced forecasting accuracy, we developed SAGE, a hybrid forecasting system that combines human and machine generated forecasts. The system provides a platform where users can interact with machine models and thus anchor their judgments on an objective benchmark. The system also aggregates human and machine forecasts weighting both for propinquity and based on assessed skill while adjusting for overconfidence. We present results from the Hybrid Forecasting Competition (HFC) - larger than comparable forecasting tournaments - including 1085 users forecasting 398 real-world forecasting problems over eight months. Our main result is that the hybrid system generated more accurate forecasts compared to a human-only baseline which had no machine generated predictions. We found that skilled forecasters who had access to machine-generated forecasts outperformed those who only viewed historical data. We also demonstrated the inclusion of machine-generated forecasts in our aggregation algorithms improved performance, both in terms of accuracy and scalability. This suggests that hybrid forecasting systems, which potentially require fewer human resources, can be a viable approach for maintaining a competitive level of accuracy over a larger number of forecasting questions.

cs.CY↗

arXiv2024

Can LLMs Capture Human Preferences?

We explore the viability of Large Language Models (LLMs), specifically OpenAI's GPT-3.5 and GPT-4, in emulating human survey respondents and eliciting preferences, with a focus on intertemporal choices. Leveraging the extensive literature on intertemporal discounting for benchmarking, we examine responses from LLMs across various languages and compare them to human responses, exploring preferences between smaller, sooner, and larger, later rewards. Our findings reveal that both GPT models demonstrate less patience than humans, with GPT-3.5 exhibiting a lexicographic preference for earlier rewards, unlike human decision-makers. Though GPT-4 does not display lexicographic preferences, its measured discount rates are still considerably larger than those found in humans. Interestingly, GPT models show greater patience in languages with weak future tense references, such as German and Mandarin, aligning with existing literature that suggests a correlation between language structure and intertemporal preferences. We demonstrate how prompting GPT to explain its decisions, a procedure we term "chain-of-thought conjoint," can mitigate, but does not eliminate, discrepancies between LLM and human responses. While directly eliciting preferences using LLMs may yield misleading results, combining chain-of-thought conjoint with topic modeling aids in hypothesis generation, enabling researchers to explore the underpinnings of preferences. Chain-of-thought conjoint provides a structured framework for marketers to use LLMs to identify potential attributes or factors that can explain preference heterogeneity across different customers and contexts.

cs.CL↗

arXiv2024

Choice Models and Permutation Invariance: Demand Estimation in Differentiated Products Markets

Choice modeling is at the core of understanding how changes to the competitive landscape affect consumer choices and reshape market equilibria. In this paper, we propose a fundamental characterization of choice functions that encompasses a wide variety of extant choice models. We demonstrate how non-parametric estimators like neural nets can easily approximate such functionals and overcome the curse of dimensionality that is inherent in the non-parametric estimation of choice functions. We demonstrate through extensive simulations that our proposed functionals can flexibly capture underlying consumer behavior in a completely data-driven fashion and outperform traditional parametric models. As demand settings often exhibit endogenous features, we extend our framework to incorporate estimation under endogenous features. Further, we also describe a formal inference procedure to construct valid confidence intervals on objects of interest like price elasticity. Finally, to assess the practical applicability of our estimator, we utilize a real-world dataset from S. Berry, Levinsohn, and Pakes (1995). Our empirical analysis confirms that the estimator generates realistic and comparable own- and cross-price elasticities that are consistent with the observations reported in the existing literature.

econ.EM↗

arXiv2023

Interpretable Anomaly Detection in Cellular Networks by Learning Concepts in Variational Autoencoders

This paper addresses the challenges of detecting anomalies in cellular networks in an interpretable way and proposes a new approach using variational autoencoders (VAEs) that learn interpretable representations of the latent space for each Key Performance Indicator (KPI) in the dataset. This enables the detection of anomalies based on reconstruction loss and Z-scores. We ensure the interpretability of the anomalies via additional information centroids (c) using the K-means algorithm to enhance representation learning. We evaluate the performance of the model by analyzing patterns in the latent dimension for specific KPIs and thereby demonstrate the interpretability and anomalies. The proposed framework offers a faster and autonomous solution for detecting anomalies in cellular networks and showcases the potential of deep learning-based algorithms in handling big data.

cs.LG↗

arXiv2023

Causal Bandits: Online Decision-Making in Endogenous Settings

The deployment of Multi-Armed Bandits (MAB) has become commonplace in many economic applications. However, regret guarantees for even state-of-the-art linear bandit algorithms (such as Optimism in the Face of Uncertainty Linear bandit (OFUL)) make strong exogeneity assumptions w.r.t. arm covariates. This assumption is very often violated in many economic contexts and using such algorithms can lead to sub-optimal decisions. Further, in social science analysis, it is also important to understand the asymptotic distribution of estimated parameters. To this end, in this paper, we consider the problem of online learning in linear stochastic contextual bandit problems with endogenous covariates. We propose an algorithm we term $ε$-BanditIV, that uses instrumental variables to correct for this bias, and prove an $\tilde{\mathcal{O}}(k\sqrt{T})$ upper bound for the expected regret of the algorithm. Further, we demonstrate the asymptotic consistency and normality of the $ε$-BanditIV estimator. We carry out extensive Monte Carlo simulations to demonstrate the performance of our algorithms compared to other methods. We show that $ε$-BanditIV significantly outperforms other existing methods in endogenous settings. Finally, we use data from real-time bidding (RTB) system to demonstrate how $ε$-BanditIV can be used to estimate the causal impact of advertising in such settings and compare its performance with other existing methods.

econ.EM↗

arXiv2021

User-friendly Comparison of Similarity Algorithms on Wikidata

While the similarity between two concept words has been evaluated and studied for decades, much less attention has been devoted to algorithms that can compute the similarity of nodes in very large knowledge graphs, like Wikidata. To facilitate investigations and head-to-head comparisons of similarity algorithms on Wikidata, we present a user-friendly interface that allows flexible computation of similarity between Qnodes in Wikidata. At present, the similarity interface supports four algorithms, based on: graph embeddings (TransE, ComplEx), text embeddings (BERT), and class-based similarity. We demonstrate the behavior of the algorithms on representative examples about semantically similar, related, and entirely unrelated entity pairs. To support anticipated applications that require efficient similarity computations, like entity linking and recommendation, we also provide a REST API that can compute most similar neighbors for any Qnode in Wikidata.

cs.AI↗

arXiv2021

KGTK: A Toolkit for Large Knowledge Graph Manipulation and Analysis

Knowledge graphs (KGs) have become the preferred technology for representing, sharing and adding knowledge to modern AI applications. While KGs have become a mainstream technology, the RDF/SPARQL-centric toolset for operating with them at scale is heterogeneous, difficult to integrate and only covers a subset of the operations that are commonly needed in data science applications. In this paper we present KGTK, a data science-centric toolkit designed to represent, create, transform, enhance and analyze KGs. KGTK represents graphs in tables and leverages popular libraries developed for data science applications, enabling a wide audience of developers to easily construct knowledge graph pipelines for their applications. We illustrate the framework with real-world scenarios where we have used KGTK to integrate and manipulate large KGs, such as Wikidata, DBpedia and ConceptNet.

cs.AI↗

arXiv2021

Sparse-Push: Communication- & Energy-Efficient Decentralized Distributed Learning over Directed & Time-Varying Graphs with non-IID Datasets

Current deep learning (DL) systems rely on a centralized computing paradigm which limits the amount of available training data, increases system latency, and adds privacy and security constraints. On-device learning, enabled by decentralized and distributed training of DL models over peer-to-peer wirelessly connected edge devices, not only alleviate the above limitations but also enable next-gen applications that need DL models to continuously interact and learn from their environment. However, this necessitates the development of novel training algorithms that train DL models over time-varying and directed peer-to-peer graph structures while minimizing the amount of communication between the devices and also being resilient to non-IID data distributions. In this work we propose, Sparse-Push, a communication efficient decentralized distributed training algorithm that supports training over peer-to-peer, directed, and time-varying graph topologies. The proposed algorithm enables 466x reduction in communication with only 1% degradation in performance when training various DL models such as ResNet-20 and VGG11 over the CIFAR-10 dataset. Further, we demonstrate how communication compression can lead to significant performance degradation in-case of non-IID datasets, and propose Skew-Compensated Sparse Push algorithm that recovers this performance drop while maintaining similar levels of communication compression.

cs.LG↗

arXiv2021

Causal Gradient Boosting: Boosted Instrumental Variable Regression

Recent advances in the literature have demonstrated that standard supervised learning algorithms are ill-suited for problems with endogenous explanatory variables. To correct for the endogeneity bias, many variants of nonparameteric instrumental variable regression methods have been developed. In this paper, we propose an alternative algorithm called boostIV that builds on the traditional gradient boosting algorithm and corrects for the endogeneity bias. The algorithm is very intuitive and resembles an iterative version of the standard 2SLS estimator. Moreover, our approach is data driven, meaning that the researcher does not have to make a stance on neither the form of the target function approximation nor the choice of instruments. We demonstrate that our estimator is consistent under mild conditions. We carry out extensive Monte Carlo simulations to demonstrate the finite sample performance of our algorithm compared to other recently developed methods. We show that boostIV is at worst on par with the existing methods and on average significantly outperforms them.

econ.EM↗

arXiv2020

Simulating the effect of weak measurements by a phase damping channel and determining different measures of bipartite correlations in nuclear magnetic resonance

Quantum discord is a measure based on local projective measurements which captures quantum correlations that may not be fully captured by entanglement. A change in the measurement process, achieved by replacing rank-one projectors with a weak positive operator-valued measure (POVM), allows one to define weak variants of quantum discord. In this work, we experimentally simulate the effect of a weak POVM on a nuclear magnetic resonance quantum information processor. The two-qubit system under investigation is part of a three-qubit system, where one of the qubits is used as an ancillary to implement the phase damping channel. The strength of the weak POVM is controlled by varying the strength of the phase damping channel. We experimentally observed two weak variants of quantum discord namely, super quantum discord and weak quantum discord, in two-qubit Werner and Bell-diagonal states. The resultant dynamics of the states is investigated as a function of the measurement strength.

quant-ph↗

arXiv2020

Experimental Detection of Non-local Correlations using a Local Measurement-Based Hierarchy on an NMR Quantum Processor

The non-local nature of the correlations possessed by quantum systems may be revealed by experimental demonstrations of the violation of Bell-type inequalities. Recent work has placed bounds on the correlations that quantum systems can possess in an actual experiment. These bounds were limited to a composite quantum system comprising of a few lower-dimensional subsystems. In a more general approach, it has been shown that fewer body correlations can reveal the non-local nature of the correlations arising from a quantum mechanical description of nature. Such tests on the correlations can be transformed to a semi-definite program (SDP). This study reports the experimental implementation of a local measurement-based hierarchy on the nuclear magnetic resonance (NMR) hardware utilizing three nuclear spins as qubits. The protocol has been experimentally tested on genuinely entangled tripartite states such as W state, GHZ state and a few graph states. In all the cases, the experimentally measured correlations were used to formulate the SDP, using linear constraints on the entries of the moment matrix. We observed that for each genuinely entangled state, the SDP failed to find a semi-definite positive moment matrix consistent with the experimental data. This implies that the observed correlations can not arise from local measurements on a separable state and are hence non-local in nature, and also confirms that the states being tested are indeed entangled.

quant-ph↗

arXiv2019

Generation and Detection of Quantum Correlations and Entanglement on a Spin-Based Quantum Information Processor

This thesis focuses on the experimental creation and detection of a variety of quantum correlations using nuclear magnetic resonance hardware. Quantum entanglement, being most common and counter-intuitive, is one of the main type considered in this thesis. Quantum correlations play a major role in achieving, the much talked, computational speedup. Creation and detection of such correlations experimentally is a major thrust area in experimental quantum information processing field. Main goals of the studies undertaken in this thesis were to come up with strategies to detect the entanglement in a 'state-independent' way and with low experimental resources. The core of all the detection protocols is based on our own developed method which enables us to measure any observable with high accuracy. The experimental protocol has been successfully implemented to detect the entanglement of random two-qubit pure states. Further, the schemes for the experimental detection as well as classification of generic and general three-qubit pure states have also been devised and implemented successfully. Quantum correlation possessed by mixed and$/$or separable states e.g. non-classical, bound-entanglement and of non-local nature were also investigated. In all the investigations, the results were verified by one or more alternative ways e.g. full quantum state tomography.

quant-ph↗

arXiv2019

Experimental Detection of Qubit-Ququart Pseudo-Bound Entanglement using Three Nuclear Spins

In this work, we experimentally created and characterized a class of qubit-ququart PPT (positive under partial transpose) entangled states using three nuclear spins on an nuclear magnetic resonance (NMR) quantum information processor. Entanglement detection and characterization for systems with a Hilbert space dimension $\ge 2 \otimes 3$ is nontrivial since there are states in such systems which are both PPT as well as entangled. The experimental detection scheme that we devised for the detection of qubit-ququart PPT entanglement was based on the measurement of three Pauli operators with high precision, and is a key ingredient of the protocol in detecting entanglement. The family of PPT-entangled states considered in the current study are incoherent mixtures of five pure states. All the five states were prepared with high fidelities and the resulting PPT entangled states were prepared with mean fidelity $\ge 0.95$. The entanglement thus detected was validated by carrying out full quantum state tomography (QST).

quant-ph↗