Abstract

In recent years, widespread internet adoption and the growth in user base of various social media platforms have led to an increase in the prolifera tion of extreme speech online. While traditional language models have demon strated proficiency in distinguishing between neutral text and non-neutral text (i.e. extreme speech), categorizing the diverse types of extreme speech presents significant challenges [1][2]. The task of extreme speech classification is par ticularly nuanced, as it requires a deep understanding of socio-cultural contexts to accurately interpret the intent of the language used by the speaker. Even hu man annotators often disagree on the appropriate classification of such content, emphasizing the complex and subjective nature of this task [3]. The use of hu man moderators also presents a scaling issue, necessitating the need for auto mated systems for extreme speech classification. The recent launch of ChatGPT has drawn global attention to the potential applications of Large Language Models (LLMs) across a diverse variety of tasks. Trained on vast and diverse corpora, and demonstrating the ability to effectively capture and encode contex tual information, LLMs emerge as highly promising tools for tackling this spe cific task of extreme speech classification. In this paper, we leverage the Indian subset of the extreme speech dataset from [3] to develop an effective classifica tion framework using LLMs. We evaluate open-source Llama models against closed-source OpenAI models, finding that while pre-trained LLMs show mod erate efficacy, fine-tuning with domain-specific data significantly enhances per formance, highlighting their adaptability to linguistic and contextual nuances. Although GPT-based models outperform Llama models in zero-shot settings, the performance gap disappears after fine-tuning.

Abstract

Modeling process, operating condition, reliability, and technological variation in transistor level design introduce significant variance in device and performance parameters. This paper presents a machine-learning-based architectural approach that enhances model performance, as a step towards developing Foundation Models tailored for circuit data, accurately capturing technology and process-induced variations in leakage power, voltage, and current. The architecture incorporates the impact of varying operating conditions, including temperatures from -55°C to 125°C and supply voltage fluctuations of ±10% on 16nm HP FinFET, and 16nm, 22nm, 32nm, and 45nm HPMGK CMOS technology nodes. By enhancing the performance of baseline machine-learning models using a chained architecture to cater to the high variance in target, our C-Arch serves as a versatile framework for circuit modeling when the data spread is high. Experimental results on current, voltage, and leakage power estimation, demonstrate average improvements of up to 93.76%, 96.17%, and 88.82% in Mean Absolute Percentage Error compared to baseline models, underscoring the computational savings and viability of Foundation Models in circuit design.

Abstract

The Facility Location Problem (FLP) is a well-studied optimization problem with applications in many real-world scenarios. Past literature has explored the solutions from different perspectives to tackle FLPs. These include investigating FLPs under objective functions such as utilitarian, egalitarian, Nash welfare, etc. Also, there is no treatment for asymmetric welfare functions around the facility. We propose a unified framework, FLIGHT, to accommodate a broad class of welfare notions. The framework undergoes rigorous theoretical analysis, and we prove some structural properties of the solution to FLP. Additionally, we provide approximation bounds, which provide insight into an interesting fact: as the number of agents arbitrarily increases, the choice of welfare notion is irrelevant. Furthermore, the paper also includes results around concentration bounds under certain distributional assumptions over the preferred locations of agents.

Abstract

This paper considers the problem of fair probabilistic binary classification with binary protected groups. The classifier assigns scores, and a practitioner predicts labels using a certain cut-off threshold based on the desired trade-off between false positives vs. false negatives. It derives these thresholds from the ROC of the classifier. The resultant classifier may be unfair to one of the two protected groups in the dataset. It is desirable that no matter what threshold the practitioner uses, the classifier should be fair to both the protected groups; that is, the ℒₚ norm between FPRs and TPRs of both the protected groups should be at most ε. We call such fairness on ROCs of both the protected attributes εₚ-Equalized ROC. Given a classifier not satisfying ε₁-Equalized ROC, we aim to design a post-processing method to transform the given (potentially unfair) classifier's output (score) to a suitable randomized yet fair classifier. That is, the resultant classifier must satisfy ε₁-Equalized ROC. First, we introduce a threshold query model on the ROC curves for each protected group. The resulting classifier is bound to face a reduction in AUC. With the proposed query model, we provide a rigorous theoretical analysis of the minimal AUC loss to achieve ε₁-Equalized ROC. To achieve this, we design a linear time algorithm, namely FROC, to transform a given classifier's output to a probabilistic classifier that satisfies ε₁-Equalized ROC. We prove that under certain theoretical conditions, FROC achieves the theoretical optimal guarantees. We also study the performance of our FROC on multiple real-world datasets with many trained classifiers.

Abstract

Abstract— Transgenic crops have become widespread but have associated health concerns. Testing for transgenic corn is slow and lengthy. In this paper, we used UV-Vis spectroscopy in combination with chemometrics and the PCA-SVM classifier to identify different types of corn samples from each other, including transgenic varieties. The presented methodology and the classifier can differentiate very well between the three different varieties of corn considered in this paper. Although there is a slight improvement in prediction accuracies compared to the state-of-the-art methods, our method is considerably more cost-effective and less time-consuming.

Abstract

Disenchanted Tales from the Old Home in Qissa: A Tale of a Lonely Ghost Abstract: The temporal fixity of the event of Independence/Partition in nationalist historiography has allowed generations of Indians to imagine the Partition and the accompanying genocide as restricted to the year and event of 1947, and yet continue to repeat its violence in an attempt to complete the task of securing religious identities to their assigned national spaces. Just as the political histories of the nations created by the Partition continue to live with the spectres of its violence through repeated wars, so do citizen-subjects spawned by the new nations live in the shadows of old identities. This paper proposes that Singh’s Qissa: A Tale of a Lonely Ghost presents the viewer with the spectre of a partition that never ended. This paper shall explore the function of memory in Anup Singh’s 2013 feature Qissa: A Tale of a Lonely Ghost in producing and de-stabilizing gender, communal and familial identities, and real and imaginary histories. The paper shall look at specific stylistic choices of the film’s form to suggest that this historic cleavage the film populates with an indeterminacy of the corporal/non-corporal body, of the tentative presence of a ghost of days past, and the dead who didn’t stay buried. The site of memory here is inhabited by fluid bodies that lie between genders, in between the past, present and the future, and ultimately in between the living and the dead. Memory here serves as an imaginary history.

Abstract

Graph Neural Network (GNN) research is rapidly advancing due to GNNs’ capacity to learn distributed representations from graph-structured data. However, centralizing large volumes of real-world graph data for GNN training is often impractical due to privacy concerns, regulatory restrictions, and commercial competition. Federated learning (FL), a distributed learning paradigm, offers a solution by preserving data privacy with collaborative model training. Despite progress in training huge vision and language models, federated learning for GNNs remains underexplored. To address this challenge, we present a novel method for federated learning on GNNs based on spectral GNNs equipped with neural ordinary differential equations (ODE) for better information capture, showing promising results across both homophilic and heterophilic graphs. Our approach effectively handles non-Independent and Identically Distributed (non-IID) data, while also achieving performance comparable to existing methods that only operate on IID data. It is designed to be privacy-preserving and bandwidth-optimized, making it suitable for real-world applications such as social network analysis, recommendation systems, and fraud detection, which often involve complex, non-IID, and heterophilic graph structures. Our results in the area of federated learning on non-IID heterophilic graphs demonstrate significant improvements while also achieving better performance on homophilic graphs. This work highlights the potential of federated learning in diverse and challenging graph settings.

Abstract

Mann ki Baat is a series of speeches made by the 14th Prime Minister of India, Narendra Modi addressing the nation. Mann ki Baat holds significant sway over the Indian population as it is broadcast directly by the Prime Minister. As such, it is possible for this stream of speeches to carry hidden political intent or to act as a symbol to show that the government is in touch with the people. For this purpose, the English transcripts of 84 speeches from this series starting from October 2014 to March 2022 were analysed. For an overarching general analysis, Natural Language Processing tools were employed and later on, the speeches were juxtaposed with important events that have occurred in the country to investigate the responses as shown in Mann ki Baat. Moreover, 25 speeches were analysed individually in connection with 3 major events, each spanning a course of multiple months to validate the observations found through the employment of NLP tools and add to them. Observations have shown that Mann ki Baat does indeed carry political intent and works to better the image the people have of the Prime Minister and the ruling government, even at the cost of bending facts at times. However, it has also come to light that Mann ki Baat also works as a highly efficient means of spreading general awareness such as the instances observed during the Covid pandemic. Keywords: Mann ki Baat, NLP, Textual Analysis, Indian Politics

Abstract

The fully entangled fraction (FEF) measures the proximity of a quantum state to maximally entangled states. FEF , in systems, is a significant benchmark for various quantum information processing protocols including teleportation. Quantum conditional entropy (QCE) on the other hand is a measure of correlation in quantum systems. Conditional entropies for quantum systems can be negative, marking a departure from conventional classical systems. The negativity of quantum conditional entropies plays a decisive role in tasks like state merging and dense coding. In the present work, we investigate the relation of these two important yardsticks. Our probe is mainly done in the ambit of states with maximally mixed marginals, with a few illustrations from other classes of quantum states. We start our study in two-qubit systems, where for the Werner states, we obtain lower bounds to its FEF when the conditional Rényi entropy is negative. We then obtain relations between FEF and QCE for two-qubit Weyl states. Moving on to two qudit states, we find a necessary and sufficient condition based on FEF, for the isotropic state to have negative conditional entropy. In two qudit systems, the relation between FEF and QCE is probed for the rank-deficient and generalized Bell diagonal states. FEF is intricately linked with k-copy nonlocality and k- copy steerability. The relations between FEF and QCE facilitates to find conditions for k- copy nonlocality and k- copy steerability based on QCE. We obtain such conditions for certain classes of states in two qubits and two qudits. Applications of the relations obtained are provided in the context of work extraction, faithful entanglement and entropic uncertainty relations.

Abstract

Linearly Transformed Spherical Distributions (LTSDs), a superset of the commonly known Linearly Transformed Cosines (LTCs) for analytic area light rendering, applied in the context of analytic area light rendering with participating media.