相关论文: AxBench: Steering LLMs? Even Simple Baselines Outp…

Sparse autoencoders (SAEs) have emerged as a promising approach in language model interpretability, offering unsupervised extraction of sparse features. For interpretability methods to succeed, they must identify abstract features across…

Steering vectors are a promising approach to aligning language model behavior at inference time. In this paper, we propose a framework to assess the limitations of steering vectors as alignment mechanisms. Using a framework of transformer…

LLMs are increasingly being used in healthcare. This promises to free physicians from drudgery, enabling better care to be delivered at scale. But the use of LLMs in this space also brings risks; for example, such models may worsen existing…

As Large Language Models become integral to software development, with substantial portions of AI-suggested code entering production, understanding their internal correctness mechanisms becomes critical for safe deployment. We apply sparse…

The rapid advancement of large language models (LLMs) has made machine-generated text increasingly difficult to distinguish from human-written text. While recent studies explore leveraging internal representations of language models to…

The standard sparse-autoencoder (SAE) interpretability protocol labels each feature from its top-activating contexts and validates by single-feature steering. We propose the pairwise matrix protocol, co-varying steering coefficient with…

Sparse autoencoders (SAEs) have emerged as a powerful technique for decomposing language model representations into interpretable features. Current interpretation methods infer feature semantics from activation patterns, but overlook that…

Multimodal Large Language Models (MLLMs) often exhibit significant modality preference, which is a tendency to favor one modality over another. Depending on the input, they may over-rely on linguistic priors relative to visual evidence, or…

The rapid advancements in Large Language Models (LLMs) have revolutionized natural language processing (NLP) and related fields. However, fine-tuning these models for specific tasks remains computationally expensive and risks degrading…

As LLMs achieved breakthroughs in general reasoning, their proficiency in specialized scientific domains reveals pronounced gaps in existing benchmarks due to data contamination, insufficient complexity, and prohibitive human labor costs.…

Sparse Autoencoders (SAEs) have emerged as a powerful paradigm for disentangling feature superposition in transformer-based architectures, enabling precise control via activation steering. However, the theoretical foundations of…

Modern AI models contain much of human knowledge, yet understanding of their internal representation of this knowledge remains elusive. Characterizing the structure and properties of this representation will lead to improvements in model…

Modern LLMs face inference efficiency challenges due to their scale. To address this, many compression methods have been proposed, such as pruning and quantization. However, the effect of compression on a model's interpretability remains…

Large language models (LLMs) demonstrate strong chain-of-thought (CoT) reasoning abilities, while smaller models (<= 3B parameters) significantly underperform on multi-step reasoning tasks. Based on empirical analyses of the Qwen-2.5 model…

RFdiffusion is a popular and well-established model for generation of protein structures. However, this generative process offers limited insight into its internal representations and how they contribute to the final protein structure.…

We study the challenge of achieving theoretically grounded feature recovery using Sparse Autoencoders (SAEs) for the interpretation of Large Language Models. Existing SAE training algorithms often lack rigorous mathematical guarantees and…

Activation steering has emerged as a promising alternative for controlling language-model behavior at inference time by modifying intermediate representations while keeping model parameters frozen. However, large-scale evaluations such as…

Fine-tuning pre-trained transformers is a powerful technique for enhancing the performance of base models on specific tasks. From early applications in models like BERT to fine-tuning Large Language Models (LLMs), this approach has been…

Sparse autoencoders (SAEs) \citep{bricken2023monosemanticity,gao2024scalingevaluatingsparseautoencoders} rely on dictionary learning to extract interpretable features from neural networks at scale in an unsupervised manner, with…

We introduce Refusal Steering, an inference-time method to exercise fine-grained control over Large Language Models refusal behaviour on politically sensitive topics without retraining. We replace fragile pattern-based refusal detection…