3 papers
cs.LG2026
Generalised Score Matching on Convex Domains
Nishanth Shetty, Saisuchith Mahajan, Chandra Sekhar Seelamantula
Score matching avoids computing the normalising constant that maximum-likelihood estimation requires. On constrained domains, its generalised variants weight the Fisher divergence…
cs.LG2025
Dale meets Langevin: A Multiplicative Denoising Diffusion Model
Nishanth Shetty, Madhava Prasath, Chandra Sekhar Seelamantula
Exponentiated gradient descent (EGD), a biologically motivated optimisation algorithm that respects Dale's law, produces log-normally distributed synaptic weights at convergence, i…
cs.LG2023
Insights into Closed-form IPM-GAN Discriminator Guidance for Diffusion Modeling
Aadithya Srikanth, Siddarth Asokan, Nishanth Shetty +1
Diffusion models are a state-of-the-art generative modeling framework that transform noise to images via Langevin sampling, guided by the score, which is the gradient of the logari…