InterrogateLLM: Zero-Resource Hallucination Detection in LLM-Generated Answersby Yakir Yehuda, Itzik Malkiel, Oren Barkan, Jonathan Weill,…
MeanCache: User-Centric Semantic Caching for LLM Web Servicesby Waris Gill, Mohamed Elidrisi, Pallavi Kalapatapu, Ammar…
ATP: Enabling Fast LLM Serving via Attention on Top Principal Keysby Yue Niu, Saurav Prakash,…
NoMAD-Attention: Efficient LLM Inference on CPUs Through Multiply-add-free Attentionby Tianyi Zhang, Jonah Wonkyu Yi, Bowen…
Self-Refinement of Language Models from External Proxy Metrics Feedbackby Keshav Ramji, Young-Suk Lee, Ramón Fernandez…
Direct Alignment of Draft Model for Speculative Decoding with Chat-Fine-Tuned LLMsby Raghavv Goel, Mukul Gagrani,…
Brain-Inspired Two-Stage Approach: Enhancing Mathematical Reasoning by Imitating Human Thought Processesby Yezeng Chen, Zui Chen,…
Improving Socratic Question Generation using Data Augmentation and Preference Optimizationby Nischal Ashok Kumar, Andrew LanFirst…
Griffin: Mixing Gated Linear Recurrences with Local Attention for Efficient Language Modelsby Soham De, Samuel…
Keeping LLMs Aligned After Fine-tuning: The Crucial Role of Prompt Templatesby Kaifeng Lyu, Haoyu Zhao,…