10 papers
Deeper is Not Always Better: Mitigating the Alignment Tax via Confident Layer Decoding
Xuanming Zhang, Sining Zhoubian, Yuxuan Chen +8
Autoregressive generation in large language models (LLMs) conventionally decodes from the final layer, assuming that deeper representations yield more reliable next-token predictio…
ChLogic: Evaluating Robustness of Logical Reasoning in Chinese Expressions
Peixian Zhou, Yuxu Chen, Chaorui Zhang +3
Large language models perform increasingly well on standardized logical reasoning benchmarks, but whether this ability remains robust beyond English is unclear. We introduce ChLogi…
SciCore-Mol: Augmenting Large Language Models with Pluggable Molecular Cognition Modules
Yuxuan Chen, Changwei Lv, Yunduo Xiao +5
Large Language Models (LLMs) are central to the one-for-all intelligent paradigm, but they face a fundamental challenge when dealing with heterogeneous scientific data such as mole…
NaviRAG: Towards Active Knowledge Navigation for Retrieval-Augmented Generation
Jihao Dai, Dingjun Wu, Yuxuan Chen +4
Retrieval-augmented generation (RAG) typically relies on a flat retrieval paradigm that maps queries directly to static, isolated text segments. This approach struggles with more c…
UniWhisper: Efficient Continual Multi-task Training for Robust Universal Audio Representation
Yuxuan Chen, Peize He, Haoyuan Yu +1
A universal audio representation should capture fine-grained speech cues and high-level semantics for environmental sounds and music in a single encoder. Existing encoders often ex…
DrugR: Optimizing Molecular Drugs through LLM-based Explicit Reasoning
Haoran Liu, Zheni Zeng, Yukun Yan +2
Molecule generation and optimization is a fundamental task in chemical domain. The rapid development of intelligent tools, especially large language models (LLMs) with powerful kno…