4 papers
BLADE: Boundary-Expanded and Layer-Adaptive Dynamic Exit for Efficient LLM Reasoning
Keshu Fu, Keqin Peng, Jun Bai +6
Large language models often improve task performance by generating long reasoning traces, but the resulting computation is frequently wasted on redundant verification and revision.…
Xetrieval: Mechanistically Explaining Dense Retrieval
Zhixin Cai, Jun Bai, Yang Liu +7
Explaining why dense retrievers assign high relevance scores remains challenging because retrieval decisions are made through opaque high-dimensional embeddings. Existing explanati…
Selecting Demonstrations for Many-Shot In-Context Learning via Gradient Matching
Jianfei Zhang, Bei Li, Jun Bai +4
In-Context Learning (ICL) empowers Large Language Models (LLMs) for rapid task adaptation without Fine-Tuning (FT), but its reliance on demonstration selection remains a critical c…
Disentangling Preference Representation and Text Generation for Efficient Individual Preference Alignment
Jianfei Zhang, Jun Bai, Bei Li +4
Aligning Large Language Models (LLMs) with general human preferences has been proved crucial in improving the interaction quality between LLMs and human. However, human values are…