Sashank J. Reddi

A Little Help Goes a Long Way: Efficient LLM Training by Leveraging Small LMs

Efficient stagewise pretraining via progressive subnetworks

Reasoning with Latent Thoughts: On the Power of Looped Transformers

Bipartite Ranking From Multiple Labels: On Loss Versus Label Aggregation

A Little Help Goes a Long Way: Efficient LLM Training by Leveraging Small LMs

On the Role of Depth and Looping for In-Context Learning with Task Diversity

Structured Preconditioners in Adaptive Optimization: A Unified Analysis