5 citations · 14 across the 12 of their papers we have counts for
13 papers
Erase-then-Delta Attention: Decoupling Erase and Write Addresses in Delta-Rule Linear Attention
Xiao Li, Chengruidong Zhang, Hao Luo +15
Delta-rule linear attention improves recurrent memory updates by correcting what is already stored at the current write address before writing new content. However, the active corr…
Blending Supervised and Reinforcement Fine-Tuning with Prefix Sampling
Zeyu Huang, Tianhao Cheng, Zihan Qiu +4
Existing LLMs-post-training techniques are broadly categorized into supervised fine-tuning (SFT) and reinforcement fine-tuning (RFT). Each paradigm presents a distinct trade-off: (…
A Controllable Examination for Long-Context Language Models
Yijun Yang, Zeyu Huang, Wenhao Zhu +4
Existing frameworks for evaluating long-context language models (LCLM) can be broadly categorized into real-world applications (e.g, document summarization) and synthetic tasks (e.…
Post-hoc Reward Calibration: A Case Study on Length Bias
Zeyu Huang, Zihan Qiu, Zili Wang +2
Reinforcement Learning from Human Feedback aligns the outputs of Large Language Models with human values and preferences. Central to this process is the reward model (RM), which tr…
Layerwise Recurrent Router for Mixture-of-Experts
Zihan Qiu, Zeyu Huang, Shuang Cheng +4
The scaling of large language models (LLMs) has revolutionized their capabilities in various tasks, yet this growth must be matched with efficient computational strategies. The Mix…
GW-MoE: Resolving Uncertainty in MoE Router with Global Workspace Theory
Haoze Wu, Zihan Qiu, Zili Wang +2
Mixture-of-Experts (MoE) has been demonstrated as an efficient method to scale up models. By dynamically and sparsely selecting activated experts, MoE can effectively reduce comput…
A Closer Look into Mixture-of-Experts in Large Language Models
Ka Man Lo, Zeyu Huang, Zihan Qiu +2
Mixture-of-experts (MoE) is gaining increasing attention due to its unique properties and remarkable performance, especially for language tasks. By sparsely activating a subset of…
Unlocking Continual Learning Abilities in Language Models
Wenyu Du, Shuang Cheng, Tongxu Luo +5
Language models (LMs) exhibit impressive performance and generalization capabilities. However, LMs struggle with the persistent challenge of catastrophic forgetting, which undermin…
Stacking Your Transformers: A Closer Look at Model Growth for Efficient LLM Pre-Training
Wenyu Du, Tongxu Luo, Zihan Qiu +5
LLMs are computationally expensive to pre-train due to their large scale. Model growth emerges as a promising approach by leveraging smaller models to accelerate the training of la…
Empirical Study on Updating Key-Value Memories in Transformer Feed-forward Layers
Zihan Qiu, Zeyu Huang, Youcheng Huang +1
The feed-forward networks (FFNs) in transformers are recognized as a group of key-value neural memories to restore abstract high-level knowledge. In this work, we conduct an empiri…
HyperMoE: Towards Better Mixture of Experts via Transferring Among Experts
Hao Zhao, Zihan Qiu, Huijia Wu +3
The Mixture of Experts (MoE) for language models has been proven effective in augmenting the capacity of models by dynamically routing each input token to a specific subset of expe…
Heterogenous Memory Augmented Neural Networks
Zihan Qiu, Zhen Liu, Shuicheng Yan +2
It has been shown that semi-parametric methods, which combine standard neural networks with non-parametric components such as external memory modules and data retrieval, are partic…
Unlocking Emergent Modularity in Large Language Models
Zihan Qiu, Zeyu Huang, Jie Fu
Modular Neural Networks (MNNs) demonstrate various advantages over monolithic models. Existing MNNs are generally : their modular architectures are pre-defined,…