Rahul G Krishnan

Assistant Professor@Department of Computer Science, University of Toronto·加拿大·OpenReview

研究方向

unsupervised learning · Inference Networks · Time-series models · Variational Autoencoders · Variational Inference · machine learning for medicine · time series modeling · Probabilistic inference · Healthcare

Rahul G Krishnan

Can we generate portable representations for clinical time series data using LLMs?

Mitigating Privacy Risk via Forget Set-Free Unlearning

Implicit Dynamical Flow Fusion for generative modeling

CausalPFN: Amortized Causal Effect Estimation via In-Context Learning

Beyond Masked and Unmasked: Discrete Diffusion Models via Partial Masking

Reliably detecting model failures in deployment without labels

Teaching LLMs How to Learn with Contextual Fine-Tuning

Diverse Prototypical Ensembles Improve Robustness to Subpopulation Shift

AutoElicit: Using Large Language Models for Expert Prior Elicitation in Predictive Modelling

Implicit Dynamical Flow Fusion (IDFF) for Generative Modeling

Sparse Training from Random Initialization: Aligning Lottery Ticket Masks using Weight Symmetry

Provable Post-Deployment Deterioration Monitoring

Blind Unlearning: Unlearning Without a Forget Set