Aditya Kumar

Only 6.5% of neuro-symbolic AI studies can be reproduced from their published artifacts, exposing a severe reproducibility crisis in the field.

Finite-time convergence rates for risk-sensitive reinforcement learning algorithms reveal that model-free approaches can achieve robust performance without complex parameter tuning.

Efficiently removing undesirable concepts from image generations without sacrificing quality could redefine how we manage biases in generative models.