相关论文: Weight Updates as Activation Shifts: A Principled …

Test-time policy adaptation for multi-turn interactions (T2PAM) is essential for aligning Large Language Models (LLMs) with dynamic user needs during inference time. However, existing paradigms commonly treat test-time adaptation as a…

We show that training a single $d$-dimensional steering vector per layer with reinforcement learning, while freezing all base weights, matches the accuracy of fully RL-tuned reasoning models on mathematical-reasoning tasks. On an 8…

Recent advancements in language models (LMs) have marked a shift toward the growing importance of post-training. Yet, post-training approaches such as supervised fine-tuning (SFT) do not guarantee the effective use of knowledge acquired…

Continual learning requires the model to learn multiple tasks sequentially. In continual learning, the model should possess the ability to maintain its performance on old tasks (stability) and the ability to adapt to new tasks continuously…

Large language models can be steered at inference time through prompting or activation interventions, but activation steering methods often underperform compared to prompt-based approaches. We propose a framework that formulates prompt…

Post-training quantization is widely employed to reduce the computational demands of neural networks. Typically, individual substructures, such as layers or blocks of layers, are quantized with the objective of minimizing quantization…

Adapting large pretrained models to diverse tasks is now routine, yet the two dominant strategies of parameter-efficient fine-tuning (PEFT) and low-rank compression are typically composed in sequence. This decoupled practice first…

The rapid expansion of large foundation models within the pre-training and fine-tuning framework has underscored that larger models often yield better results. However, the scaling up of large foundation models has led to soaring costs in…

Post-training quantization (PTQ) methods for large language models rely on heuristics that implicitly estimate which weight channels most strongly influence model behavior. Two dominant paradigms have emerged: activation-aware methods such…

Recent progress in motion forecasting has been substantially driven by self-supervised pre-training. However, adapting pre-trained models for specific downstream tasks, especially motion prediction, through extensive fine-tuning is often…

State-of-the-art pretrained NLP models contain a hundred million to trillion parameters. Adapters provide a parameter-efficient alternative for the full finetuning in which we can only finetune lightweight neural network layers on top of…

Recently, the scale of transformers has grown rapidly, which introduces considerable challenges in terms of training overhead and inference efficiency in the scope of task adaptation. Existing works, namely Parameter-Efficient Fine-Tuning…

Aligning Large Language Models (LLMs) and Large Multimodal Models (LMMs) to human preferences is a central challenge in improving the quality of the models' generative outputs for real-world applications. A common approach is to use reward…

Large language models (LLMs) have shown impressive abilities in leveraging pretrained knowledge through prompting, but they often struggle with unseen tasks, particularly in data-scarce scenarios. While cross-task in-context learning offers…

Pretrained Foundation Models (PFMs) have transformed numerous applications by enabling efficient adaptation to customized tasks. Parameter-Efficient Fine-Tuning (PEFT) has emerged as a resource-efficient alternative to full fine-tuning,…

Steering, or direct manipulation of internal activations to guide LLM responses toward specific semantic concepts, is emerging as a promising avenue for both understanding how semantic concepts are stored within LLMs and advancing LLM…

The rapid development of parameter-efficient fine-tuning methods has noticeably improved the efficiency of adapting large language models. Among these, LoRA has gained widespread popularity due to its strong balance of effectiveness and…

Adapters have been positioned as a parameter-efficient fine-tuning (PEFT) approach, whereby a minimal number of parameters are added to the model and fine-tuned. However, adapters have not been sufficiently analyzed to understand if PEFT…

Popular PEFT methods reduce trainable parameter count for fine-tuning by parameterizing new low-rank or sparse trainable weights in parallel to the frozen pre-trained weights $W$. However, these weights are trained from scratch, and there…

Instruction tuning has become an important step for finetuning pretrained language models to better follow human instructions and generalize on various tasks. Nowadays, pretrained language models become increasingly larger, and full…