PublicationEfficient Knowledge Distillation for LLMs: Offline Top-K Logits and a Fused Chunked KL LossAug 4, 2026 · arXiv (Cornell University) · Citations: 4
PublicationQuantization-Aware Healing: A Practical Recipe for Recovering Compressed, 4-Bit LLMsAug 21, 2026 · arXiv (Cornell University)