10 papers
Cubit: Token Mixer with Kernel Ridge Regression
Chuanyang Zheng, Jiankai Sun, Yihang Gao +6
Since its introduction in 2017, the Transformer has become one of the most widely adopted architectures in modern deep learning. Despite extensive efforts to improve positional enc…
UniPool: A Globally Shared Expert Pool for Mixture-of-Experts
Minbin Huang, Han Shi, Chuanyang Zheng +5
Modern Mixture-of-Experts (MoE) architectures allocate expert capacity through a rigid per-layer rule: each transformer layer owns a separate expert set. This convention couples de…
ATTS: Asynchronous Test-Time Scaling via Conformal Prediction
Jing Xiong, Qiujiang Chen, Fanghua Ye +11
Large language models (LLMs) benefit from test-time scaling but are often hampered by high inference latency. Speculative decoding is a natural way to accelerate the scaling proces…
Understanding the Mixture-of-Experts with Nadaraya-Watson Kernel
Chuanyang Zheng, Jiankai Sun, Yihang Gao +13
Mixture-of-Experts (MoE) has become a cornerstone in recent state-of-the-art large language models (LLMs). Traditionally, MoE relies on as the router score funct…
Answer-Consistent Chain-of-thought Reinforcement Learning For Multi-modal Large Langauge Models
Minbin Huang, Runhui Huang, Chuanyang Zheng +4
Recent advances in large language models (LLMs) have demonstrated that reinforcement learning with verifiable rewards (RLVR) can significantly enhance reasoning abilities by direct…
UNComp: Can Matrix Entropy Uncover Sparsity? -- A Compressor Design from an Uncertainty-Aware Perspective
Jing Xiong, Jianghan Shen, Fanghua Ye +9
Deploying large language models (LLMs) for long-context inference remains challenging due to their substantial memory and computational demands. While techniques such as Key-Value…