Ashmita Chakraborty | Searcharxiv

arXiv2026

Dimming of Photon Ring due to Photon-Axion Conversion around Kerr Black Holes

We investigate photon-axion conversion in the vicinity of rotating Kerr black holes where strong gravity traps photons on near-circular trajectories, effectively enhancing the path length. We explore the observable signatures of such a conversion near the photon region. The process, driven by ambient magnetic fields, is significantly more efficient around supermassive black holes such as M87*, since the luminosity of photons increases with the mass of the BH. By numerically evaluating photon path lengths (on which the conversion depends), we analyze how key parameters-photon frequency, axion mass, photon-axion coupling, magnetic field strength, plasma density, and black hole spin-affect the conversion probability and the resultant dimming of photon spectral luminosity. We find that the conversion is most efficient at high frequencies (X-rays and gamma rays), while the frequency window associated with efficient conversion widens with an increase in the photon-axion coupling and a decrease in the electron density and the axion mass. The magnitude of dimming of the photon spectral luminosity depends primarily on the magnetic field, the photon-axion coupling and the BH spin. Our study reveals that rotating black holes generally exhibit enhanced dimming compared to static ones. Thus, if future telescopes achieving a resolution $\sim 10^{-5}$ arcsec in the X-ray/gamma-ray band detect a dimming of the photon spectral luminosity, then they can provide interesting constraints on the axion mass and its coupling with photons.

gr-qc↗

arXiv2025

Conversation AI Dialog for Medicare powered by Finetuning and Retrieval Augmented Generation

Large language models (LLMs) have shown impressive capabilities in natural language processing tasks, including dialogue generation. This research aims to conduct a novel comparative analysis of two prominent techniques, fine-tuning with LoRA (Low-Rank Adaptation) and the Retrieval-Augmented Generation (RAG) framework, in the context of doctor-patient chat conversations with multiple datasets of mixed medical domains. The analysis involves three state-of-the-art models: Llama-2, GPT, and the LSTM model. Employing real-world doctor-patient dialogues, we comprehensively evaluate the performance of models, assessing key metrics such as language quality (perplexity, BLEU score), factual accuracy (fact-checking against medical knowledge bases), adherence to medical guidelines, and overall human judgments (coherence, empathy, safety). The findings provide insights into the strengths and limitations of each approach, shedding light on their suitability for healthcare applications. Furthermore, the research investigates the robustness of the models in handling diverse patient queries, ranging from general health inquiries to specific medical conditions. The impact of domain-specific knowledge integration is also explored, highlighting the potential for enhancing LLM performance through targeted data augmentation and retrieval strategies.

cs.CL↗