Related papers: MSign: An Optimizer Preventing Training Instabilit…
Machine unlearning in foundation models (e.g., language and vision transformers) is essential for privacy and safety; however, existing approaches are unstable and unreliable. A widely used strategy, the gradient difference method, applies…
Scaling deep reinforcement learning networks is challenging and often results in degraded performance, yet the root causes of this failure mode remain poorly understood. Several recent works have proposed mechanisms to address this, but…
Recent developments in large language models (LLMs) have introduced new requirements for efficient and robust training. As LLM clusters scale, node failures, lengthy recoveries, and bulky checkpoints erode efficiency. Infrequent…
Large Language Models (LLMs) achieve competitive performance across diverse natural language processing (NLP) tasks, yet pretraining is computationally demanding, making optimizer efficiency an important practical consideration. Muon…
Recent research has highlighted the importance of dataset size in scaling language models. However, large language models (LLMs) are notoriously token-hungry during pre-training, and high-quality text data on the web is approaching its…
High sensitivity of neural networks against malicious perturbations on inputs causes security concerns. To take a steady step towards robust classifiers, we aim to create neural network models provably defended from perturbations. Prior…
Neural collapse ($\mathcal{NC}$) is a phenomenon observed in classification tasks where top-layer representations collapse into their class means, which become equinorm, equiangular and aligned with the classifiers. These behaviours --…
Large Language Models (LLMs) have benefited enormously from scaling, yet these gains are bounded by five fundamental limitations: (1) hallucination, (2) context compression, (3) reasoning degradation, (4) retrieval fragility, and (5)…
Fine-tuning pretrained models has become a standard approach to adapting pretrained knowledge to improve the accuracy on new sparse, imbalance datasets. However, issues arise when optimization falls into a collapsed state, where the model…
Training extremely large language models (LLMs) with billions of parameters is a computationally intensive task that pushes the limits of current data parallel training systems. While techniques like ZeRO++ have enabled efficient…
The recent success of Large Language Models (LLMs) has been predominantly driven by curating the training dataset composition, scaling of model architectures and dataset sizes and advancements in pretraining objectives, leaving tokenizer…
Pre-trained language models (PLMs) have achieved remarkable success in NLP tasks. Despite the great success, mainstream solutions largely follow the pre-training then finetuning paradigm, which brings in both high deployment costs and low…
Large Language Models (LLMs) are highly vulnerable to input perturbations, as even a small prompt change may result in a substantially different output. Existing methods to enhance LLM robustness are primarily focused on perturbed data…
Recent state-of-the-art language models utilize a two-phase training procedure comprised of (i) unsupervised pre-training on unlabeled text, and (ii) fine-tuning for a specific supervised task. More recently, many studies have been focused…
Large Language Models (LLMs) are widely deployed in real-world applications, yet little is known about their training dynamics at the token level. Evaluation typically relies on aggregated training loss, measured at the batch level, which…
This work investigates the optimization instability of deep neural networks from a less-explored yet insightful perspective: the emergence and amplification of singularities in the parametric space. Our analysis reveals that parametric…
Large language models(LLMs) containing tens of billions of parameters (or even more) have demonstrated impressive capabilities in various NLP tasks. However, substantial model size poses challenges to training, inference, and deployment so…
This study investigates the impact of localized updates to large language models (LLMs), specifically in the context of knowledge editing - a task aimed at incorporating or modifying specific facts without altering broader model…
The paradigm of Multimodal Large Language Models (MLLMs) offers a promising blueprint for advancing the electromagnetic (EM) domain. However, prevailing approaches often deviate from the native MLLM paradigm, instead using task-specific or…
Fueled by their remarkable ability to tackle diverse tasks across multiple domains, large language models (LLMs) have grown at an unprecedented rate, with some recent models containing trillions of parameters. This growth is accompanied by…