David Daniel Cox

EvoScale: Evolutionary Test-Time Scaling for Software Engineering

Rank-efficient Mixture of Experts for LLM Finetuning

Self-MoE: Towards Compositional Large Language Models with Self-Specialized Experts

Activated LoRA: Fine-tuned LLMs for Intrinsics

Satori: Reinforcement Learning with Chain-of-Action-Thought Enhances LLM Reasoning via Autoregressive Search

Power Scheduler: A Batch Size and Token Number Agnostic Learning Rate Scheduler