14 papers
NeUQI: Near-Optimal Uniform Quantization Parameter Initialization for Low-Bit LLMs
Li Lin, Xinyu Hu, Xiaojun Wan
Large language models (LLMs) achieve impressive performance across domains but face significant challenges when deployed on consumer-grade GPUs or personal devices such as laptops,…
LEDOM: Reverse Language Model
Xunjian Yin, Sitao Cheng, Yuxi Xie +6
Autoregressive language models are trained exclusively left-to-right. We explore the complementary factorization, training right-to-left at scale, and ask what reasoning patterns e…
CE-RM: A Pointwise Generative Reward Model Optimized via Two-Stage Rollout and Unified Criteria
Xinyu Hu, Yancheng He, Weixun Wang +6
Automatic evaluation is crucial yet challenging for open-ended natural language generation, especially when rule-based metrics are infeasible. Compared with traditional methods, th…
LoaQ: Layer-wise Output Approximation Quantization
Li Lin, Xiaojun Wan
A natural and intuitive idea in model quantization is to approximate each component's quantized output to match its original. Motivated by this idea, most layer-wise post-training…
Who Writes What: Unveiling the Impact of Author Roles on AI-generated Text Detection
Jiatao Li, Xiaojun Wan
The rise of Large Language Models (LLMs) necessitates accurate AI-generated text detection. However, current approaches largely overlook the influence of author characteristics. We…
SCOPE: Intrinsic Semantic Space Control for Mitigating Copyright Infringement in LLMs
Zhenliang Zhang, Xinyu Hu, Xiaojun Wan
Large language models sometimes inadvertently reproduce passages that are copyrighted, exposing downstream applications to legal risk. Most existing studies for inference-time defe…