English
Related papers

Related papers: OASIS: Online Activation Subspace Learning for Mem…

200 papers

Recent vision-language-action (VLA) models and world action models (WAMs) advance robotic manipulation by enriching intermediate representations with auxiliary spatial features or future visual-state prediction. However, these…

Robotics · Computer Science 2026-05-26 Xinzhe Chen , Sihua Ren , Liqi Huang , Haowen Sun , Mingyang Li , Xingyu Chen , Zeyang Liu , Xuguang Lan

Large Language Models (LLMs) face a crucial challenge from fixed context windows and inadequate memory management, leading to a severe shortage of long-term memory capabilities and limited personalization in the interactive experience with…

Artificial Intelligence · Computer Science 2025-06-10 Jiazheng Kang , Mingming Ji , Zhe Zhao , Ting Bai

Training large language models (LLMs) is often bottlenecked by extreme memory demands, with optimizer states dominating the footprint. Recent works mitigates this cost by projecting gradients into low-dimensional subspaces using…

Machine Learning · Computer Science 2025-10-03 Sahar Rajabi , Nayeema Nonta , Samanvay Vajpayee , Sirisha Rambhatla

Fine-tuning large language models (LLMs) for specialized domains often necessitates a trade-off between acquiring domain expertise and retaining general reasoning capabilities, a phenomenon known as catastrophic forgetting. Existing…

Machine Learning · Computer Science 2026-02-09 Xiyang Zhang , Yuanhe Tian , Hongzhi Wang , Yan Song

There has been a growing interest in enhancing rule-based agent-based models (ABMs) for social media platforms (i.e., X, Reddit) with more realistic large language model (LLM) agents, thereby allowing for a more nuanced study of complex…

Training large language models (LLMs) relies on adaptive optimizers such as Adam, which introduce extra operations and require significantly more memory to maintain first- and second-order moments than SGD. While recent works such as…

Machine Learning · Computer Science 2026-05-22 Athanasios Glentis , Jiaxiang Li , Andi Han , Mingyi Hong

Training machine learning models on mobile devices has the potential of improving both privacy and accuracy of the models. However, one of the major obstacles to achieving this goal is the memory limitation of mobile devices. Reducing…

Audio and Speech Processing · Electrical Eng. & Systems 2020-01-27 Mary Gooneratne , Khe Chai Sim , Petr Zadrazil , Andreas Kabel , Françoise Beaufays , Giovanni Motta

On-device learning is essential for personalization, privacy, and long-term adaptation in resource-constrained environments. Achieving this requires efficient learning, both fine-tuning existing models and continually acquiring new tasks…

Machine Learning · Computer Science 2026-03-18 Marco Paul E. Apolinario , Kaushik Roy

Large language models (LLMs) often suffer from catastrophic forgetting in continual learning (CL) scenarios, where performance on previously learned tasks degrades severely while training on sequentially arriving tasks. Although pioneering…

Machine Learning · Computer Science 2025-10-17 Zhiyi Wan , Wanrou Du , Liang Li , Miao Pan , Xiaoqi Qin

The expansion of machine learning into dynamic environments presents challenges in handling open-world problems where label shift, covariate shift, and unknown classes emerge. Post-training methods have been explored to address these…

Machine Learning · Computer Science 2025-08-26 Miru Kim , Mugon Joe , Minhae Kwon

Training Large Language Models (LLMs) presents significant memory challenges, predominantly due to the growing size of weights and optimizer states. Common memory-reduction approaches, such as low-rank adaptation (LoRA), add a trainable…

Machine Learning · Computer Science 2024-06-04 Jiawei Zhao , Zhenyu Zhang , Beidi Chen , Zhangyang Wang , Anima Anandkumar , Yuandong Tian

Existing low-rank adaptation (LoRA) methods face challenges on sparse large language models (LLMs) due to the inability to maintain sparsity. Recent works introduced methods that maintain sparsity by augmenting LoRA techniques with…

Computation and Language · Computer Science 2025-01-16 Yuxuan Hu , Jing Zhang , Xiaodong Chen , Zhe Zhao , Cuiping Li , Hong Chen

Benefiting from massive corpora and advanced hardware, large language models (LLMs) exhibit remarkable capabilities in language understanding and generation. However, their performance degrades in scenarios where multiple tasks are…

Computation and Language · Computer Science 2023-10-24 Xiao Wang , Tianze Chen , Qiming Ge , Han Xia , Rong Bao , Rui Zheng , Qi Zhang , Tao Gui , Xuanjing Huang

The proliferation of spectroscopic data across various scientific and engineering fields necessitates automated processing. We introduce OASIS (Omni-purpose Analysis of Spectra via Intelligent Systems), a machine learning (ML) framework for…

Machine Learning · Computer Science 2025-09-16 Chris Young , Juejing Liu , Marie L. Mortensen , Yifu Feng , Elizabeth Li , Zheming Wang , Xiaofeng Guo , Kevin M. Rosso , Xin Zhang

Modern optimizers like Adam and Muon are central to training large language models, but their reliance on first- and second-order momenta introduces significant memory overhead, which constrains scalability and computational efficiency. In…

Machine Learning · Computer Science 2026-03-02 Zhengbo Wang , Jian Liang , Ran He , Zilei Wang , Tieniu Tan

The recent paradigm shift to large-scale foundation models has brought about a new era for deep learning that, while has found great success in practice, has also been plagued by prohibitively expensive costs in terms of high memory…

Machine Learning · Computer Science 2025-05-21 Stephen Zhang , Vardan Papyan

We propose OASIS, an outlier- and sink-aware technique built on inter-layer null signaling. As AttnResidual architectures introduce an additional depth-wise normalization channel, they improve inter-layer routing flexibility but also…

There has been an explosion of interest in designing high-performance Transformers. While Transformers have delivered significant performance improvements, training such networks is extremely memory intensive owing to storing all…

Computer Vision and Pattern Recognition · Computer Science 2022-08-30 Zizheng Pan , Peng Chen , Haoyu He , Jing Liu , Jianfei Cai , Bohan Zhuang

Owing to the huge success of generative artificial intelligence (AI), large language models (LLMs) have emerged as a core subclass, underpinning applications such as question answering, text generation, and code completion. While…

Distributed, Parallel, and Cluster Computing · Computer Science 2026-02-03 Yong-Cheng Liaw , Shuo-Han Chen

The Active Subspace (AS) method is a widely used technique for identifying the most influential directions in high-dimensional input spaces that affect the output of a computational model. The standard AS algorithm requires a sufficient…

Numerical Analysis · Mathematics 2025-10-24 Fabio Nobile , Matteo Raviola , Raul Tempone