collaborators

6 papers

cs.LG2026

Two Dimensions Govern Agnostic Multiclass Transductive Learning

Pahan Dewasurendra

In transductive classification, an adversary fixes a labeled population, one label is hidden uniformly, and the learner sees all remaining labels. For binary classes, agnostic tran…

cs.GT2026

Self-Bounding Regret Matching+ in Potential Games and Product-Simplex Optimization

Pahan Dewasurendra, Subhashini Jayawardhana

Regret matching+ (RM+) is parameter free, scale invariant, and central to large game solving, but its only general individual-regret guarantee grows as . A recent ICLR re…

cs.CC2026

Convex Networks Remain Hard to Certify: Dimension-Accuracy Barriers for Lipschitz Constants

Pahan Dewasurendra, Subhashini Jayawardhana

Input-convex neural networks permit globally tractable minimization over their inputs, so one might expect their global regularity to be tractable in low input dimension. We prove…

cs.LG2026

Multiscale Reward Hedging from Correct Demonstrations

Pahan Dewasurendra

Learning from correct demonstrations is harder than supervised learning when many answers are correct: after predicting, the learner sees one valid answer but not whether its own a…

cs.LG2026

Dirichlet Follow-the-Leader Closes the Gap in Simultaneous Multiclass U-Calibration

Pahan Dewasurendra

Can one forecaster attain the optimal regret rate for every bounded proper loss and also adapt to every smooth proper loss? Recent work answered this up to a dimension gap. Its sel…

cs.LG2024

Unsupervised Replay Strategies for Continual Learning with Limited Data

Anthony Bazhenov, Pahan Dewasurendra, Giri P. Krishnan +1

Artificial neural networks (ANNs) show limited performance with scarce or imbalanced training data and face challenges with continuous learning, such as forgetting previously learn…