200 citations · 550 across the 28 of their papers we have counts for
28 papers
FlexiGPT: Pruning and Extending Large Language Models with Low-Rank Weight Sharing
James Seale Smith, Chi-Heng Lin, Shikhar Tuli +5
The rapid proliferation of large language models (LLMs) in natural language processing (NLP) has created a critical need for techniques that enable efficient deployment on memory-c…
DISP-LLM: Dimension-Independent Structural Pruning for Large Language Models
Shangqian Gao, Chi-Heng Lin, Ting Hua +4
Large Language Models (LLMs) have achieved remarkable success in various natural language processing tasks, including language modeling, understanding, and generation. However, the…
MoDeGPT: Modular Decomposition for Large Language Model Compression
Chi-Heng Lin, Shangqian Gao, James Seale Smith +5
Large Language Models (LLMs) have reshaped the landscape of artificial intelligence by demonstrating exceptional performance across various tasks. However, substantial computationa…
DynaMo: Accelerating Language Model Inference with Dynamic Multi-Token Sampling
Shikhar Tuli, Chi-Heng Lin, Yen-Chang Hsu +3
Traditional language models operate autoregressively, i.e., they predict one token at a time. Rapid explosion in model sizes has resulted in high inference times. In this work, we…
Token Fusion: Bridging the Gap between Token Pruning and Token Merging
Minchul Kim, Shangqian Gao, Yen-Chang Hsu +2
Vision Transformers (ViTs) have emerged as powerful backbones in computer vision, outperforming many traditional CNNs. However, their computational overhead, largely attributed to…
Continual Diffusion with STAMINA: STack-And-Mask INcremental Adapters
James Seale Smith, Yen-Chang Hsu, Zsolt Kira +2
Recent work has demonstrated a remarkable ability to customize text-to-image diffusion models to multiple, fine-grained concepts in a sequential (i.e., continual) manner while only…
Training Energy-Based Normalizing Flow with Score-Matching Objectives
Chen-Hao Chao, Wei-Fang Sun, Yen-Chang Hsu +2
In this paper, we establish a connection between the parameterization of flow-based and energy-based generative models, and present a new flow-based modeling approach called energy…
Continual Diffusion: Continual Customization of Text-to-Image Diffusion with C-LoRA
James Seale Smith, Yen-Chang Hsu, Lingyu Zhang +4
Recent works demonstrate a remarkable ability to customize text-to-image diffusion models while only providing a few example images. What happens if you try to customize such model…
Numerical Optimizations for Weighted Low-rank Estimation on Language Model
Ting Hua, Yen-Chang Hsu, Felicity Wang +3
Singular value decomposition (SVD) is one of the most popular compression methods that approximate a target matrix with smaller matrices. However, standard SVD treats the parameter…
Language model compression with weighted low-rank factorization
Yen-Chang Hsu, Ting Hua, Sungen Chang +3
Factorizing a large matrix into small matrices is a popular strategy for model compression. Singular value decomposition (SVD) plays a vital role in this compression strategy, appr…
A Closer Look at Knowledge Distillation with Features, Logits, and Gradients
Yen-Chang Hsu, James Smith, Yilin Shen +2
Knowledge distillation (KD) is a substantial strategy for transferring learned knowledge from one neural network model to another. A vast number of methods have been developed for…
A Closer Look at Rehearsal-Free Continual Learning
James Seale Smith, Junjiao Tian, Shaunak Halbe +2
Continual learning is a setting where machine learning models learn novel concepts from continuously shifting training data, while simultaneously avoiding degradation of knowledge…
Hyperparameter-free Continuous Learning for Domain Classification in Natural Language Understanding
Ting Hua, Yilin Shen, Changsheng Zhao +2
Domain classification is the fundamental task in natural language understanding (NLU), which often requires fast accommodation to new emerging domains. This constraint makes it imp…
Exploring Covariate and Concept Shift for Detection and Calibration of Out-of-Distribution Data
Junjiao Tian, Yen-Change Hsu, Yilin Shen +2
Moving beyond testing on in-distribution data works on Out-of-Distribution (OOD) detection have recently increased in popularity. A recent attempt to categorize OOD data introduces…
A Geometric Perspective towards Neural Calibration via Sensitivity Decomposition
Junjiao Tian, Dylan Yung, Yen-Chang Hsu +1
It is well known that vision classification models suffer from poor calibration in the face of data distribution shifts. In this paper, we take a geometric approach to this problem…
Enhancing the Generalization for Intent Classification and Out-of-Domain Detection in SLU
Yilin Shen, Yen-Chang Hsu, Avik Ray +1
Intent classification is a major task in spoken language understanding (SLU). Since most models are built with pre-collected in-domain (IND) training utterances, their ability to d…
Always Be Dreaming: A New Approach for Data-Free Class-Incremental Learning
James Smith, Yen-Chang Hsu, Jonathan Balloch +3
Modern computer vision applications suffer from catastrophic forgetting when incrementally learning new concepts over time. The most successful approaches to alleviate this forgett…
Memory-Efficient Semi-Supervised Continual Learning: The World is its Own Replay Buffer
James Smith, Jonathan Balloch, Yen-Chang Hsu +1
Rehearsal is a critical component for class-incremental continual learning, yet it requires a substantial memory budget. Our work investigates whether we can significantly reduce t…
Posterior Re-calibration for Imbalanced Datasets
Junjiao Tian, Yen-Cheng Liu, Nathan Glaser +2
Neural Networks can perform poorly when the training label distribution is heavily imbalanced, as well as when the testing data differs from the training distribution. In order to…
Generalized ODIN: Detecting Out-of-distribution Image without Learning from Out-of-distribution Data
Yen-Chang Hsu, Yilin Shen, Hongxia Jin +1
Deep neural networks have attained remarkable performance when applied to data that comes from the same distribution as that of the training set, but can significantly degrade othe…
Multi-class Classification without Multi-class Labels
Yen-Chang Hsu, Zhaoyang Lv, Joel Schlosser +2
This work presents a new strategy for multi-class classification that requires no class-specific labels, but instead leverages pairwise similarity between examples, which is a weak…
Re-evaluating Continual Learning Scenarios: A Categorization and Case for Strong Baselines
Yen-Chang Hsu, Yen-Cheng Liu, Anita Ramasamy +1
Continual learning has received a great deal of attention recently with several approaches being proposed. However, evaluations involve a diverse set of scenarios making meaningful…
A probabilistic constrained clustering for transfer learning and image category discovery
Yen-Chang Hsu, Zhaoyang Lv, Joel Schlosser +2
Neural network-based clustering has recently gained popularity, and in particular a constrained clustering formulation has been proposed to perform transfer learning and image cate…
Learning to Cluster for Proposal-Free Instance Segmentation
Yen-Chang Hsu, Zheng Xu, Zsolt Kira +1
This work proposed a novel learning objective to train a deep neural network to perform end-to-end image pixel clustering. We applied the approach to instance segmentation, which i…