Nived Rajaraman

From Markov to Laplace: How Mamba In-Context Learns Markov Chains

Interactive Learning of Single-Index Models via Stochastic Gradient Descent

SPECS: Faster Test-Time Scaling through Speculative Drafts and Dynamic Switching

What One Cannot, Two Can: Two-Layer Transformers Provably Represent Induction Heads on Any-Order Markov Chains

Scaling Test-Time Compute Without Verification or RL is Suboptimal