arXiv subjects

Arpita Kundu

Publications and source records attributed to Arpita Kundu.

6 recordsLinked to original sources

arXiv2026

Generative AI Alignment with Hinduism's Theological Plurality and Sacred Representation

Generative AI systems are increasingly used to answer personal questions and mediate everyday practices, including religion. However, existing discussions around AI alignment and ethics have largely centered secular, Western, and Abrahamic assumptions about religion, offering limited attention to other faith-based traditions. In this paper, we examine how Hindu users engage with generative AI systems in relation to their religious knowledge, belief, and practice. Drawing on 15 semi-structured interviews with Bangladeshi Hindu participants, we analyze how users interpret AI-generated religious representations, scriptural explanations, devotional interactions, and synthetic religious media. We found that AI can be both accessible and ethically troubling. While AI supported scriptural inquiry, devotional visualization, and religious storytelling, our study also identified concerns about theological flattening, cultural misrepresentation, devotional manipulation, and the simulation of sacred presence and authority. We conclude by arguing that religious alignment in generative AI requires interpretive alignment: systems that disclose their limits, preserve plurality, and avoid simulating sacred authority and sycophantic personalization.

cs.AI↗

arXiv2026

Identification of L-functions in the extended Selberg class by preimages of finite sets

In 2023, Li, Du, Yi proved a uniqueness theorem for L functions in the extended Selberg class under the assumptions of positive degree, a shared functional equation, and the sharing of three complex values. This was later strengthened by the present authors, who showed that sharing an arbitrary finite set of complex values, counted with multiplicities, still forces equality of the two L functions, again under the assumption that they satisfy the same functional equation. In this paper, we significantly improve all these results. We completely remove the requirement that the two L functions satisfy the same functional equation, yet we still obtain the same strong uniqueness conclusion under far weaker hypotheses. As a major consequence, we prove that every polynomial with distinct zeros is a strong uniqueness polynomial for L functions.

math.CV↗

arXiv2025

V-SAT: Video Subtitle Annotation Tool

The surge of audiovisual content on streaming platforms and social media has heightened the demand for accurate and accessible subtitles. However, existing subtitle generation methods primarily speech-based transcription or OCR-based extraction suffer from several shortcomings, including poor synchronization, incorrect or harmful text, inconsistent formatting, inappropriate reading speeds, and the inability to adapt to dynamic audio-visual contexts. Current approaches often address isolated issues, leaving post-editing as a labor-intensive and time-consuming process. In this paper, we introduce V-SAT (Video Subtitle Annotation Tool), a unified framework that automatically detects and corrects a wide range of subtitle quality issues. By combining Large Language Models(LLMs), Vision-Language Models (VLMs), Image Processing, and Automatic Speech Recognition (ASR), V-SAT leverages contextual cues from both audio and video. Subtitle quality improved, with the SUBER score reduced from 9.6 to 3.54 after resolving all language mode issues and F1-scores of ~0.80 for image mode issues. Human-in-the-loop validation ensures high-quality results, providing the first comprehensive solution for robust subtitle annotation.

cs.LG↗

arXiv2021

Uniqueness of $L$ function with special class of meromorphic function under restricted sharing of sets

The purpose of the paper is to rectify a series of errors occurred in [2], [17], [20] for a particular situation. To get a fruitful solution and to overcome the issue, we introduce a new form of set sharing namely restricted set sharing, which is stronger than the usual one. We manipulate the newly introduced notion in this specific section of literature to resolve all the complications. Not only that we have subtly used the same sharing form to a well known unique range set [3] to settle a long time unsolved problem.

math.CV↗

arXiv2020

Uniqueness and two shared set problems of L-Function and certain class of meromorphic function

Starting with a question of Yuan-Li-Yi [Value distribution of L-functions and uniqueness questions of F. Gross, Lithuanian Math. J., 58(2)(2018), 249-262] we have studied the uniqueness of a meromorphic function f and an L-function L sharing two finite sets. At the time of execution of our work, we have pointed out a serious lacuna in the proof of a recent result of a of Sahoo-Halder [ Some results on L-functions related to sharing two finite sets, Comput. Methods Funct. Theo., 19(2019), 601-612] which makes most of the part of the Sahoo-Halder's paper under question. In context of our choice of sets, we have rectified Sahoo-Halder's result in a convenient manner.

math.NT↗

arXiv2020

Weighted sharing and uniqueness of L-Function with certain class of meromorphic function

The purpose of the paper is to study the uniqueness problem of a $L$ function in the Selberg class sharing one or two sets with an arbitrary meromorphic function having finite poles. We manipulate the notion of weighted sharing of sets to improve one result of Yuan-Li-Yi [Value distribution of $L$-functions and uniqueness questions of F. Gross, Lithuanian Math. J., $\mathbf{58(2)}$(2018), 249-262]. More importantly, we have pointed out a number of gaps in all the results of Sahoo-Halder [Results on $L$ functions and certain uniqueness question of Gross, Lithuanian Math. J., $\mathbf{60(1)}$(2020), 80-91] which actually makes the validity of the same paper under question. As an attempt to rectify the results of Sahoo-Halder we have presented the accurate forms and proof of the results in a compact and convenient manner.

math.NT↗