相关论文: Endogenous Resistance to Activation Steering in La…
A popular approach to post-training control of large language models (LLMs) is the steering of intermediate latent representations. Namely, identify a well-chosen direction depending on the task at hand and perturbs representations along…
Emergent misalignment can arise when a language model is fine-tuned on a narrowly scoped supervised objective: the model learns the target behavior, yet also develops undesirable out-of-domain behaviors. We investigate a mechanistic…
Large Language Models (LLMs) encode factual knowledge within hidden parametric spaces that are difficult to inspect or control. While Sparse Autoencoders (SAEs) can decompose hidden activations into more fine-grained, interpretable…
Altruism is fundamental to human societies, fostering cooperation and social cohesion. Recent studies suggest that large language models (LLMs) can display human-like prosocial behavior, but the internal computations that produce such…
We propose affine concept editing (ACE) as an approach for steering language models' behavior by intervening directly in activations. We begin with an affine decomposition of model activation vectors and show that prior methods for steering…
Inference-time scaling has attracted much attention which significantly enhance the performance of Large Language Models (LLMs) in complex reasoning tasks by increasing the length of Chain-of-Thought. These longer intermediate reasoning…
Activation steering has emerged as a promising approach for efficiently adapting large language models (LLMs) to downstream behaviors. However, most existing steering methods rely on a single static direction per task or concept, making…
Mitigating explicit and implicit biases in Large Language Models (LLMs) has become a critical focus in the field of natural language processing. However, many current methodologies evaluate scenarios in isolation, without considering the…
Steering language models (LMs) by modifying internal activations is a popular approach for controlling text generation. Unsupervised dictionary learning methods, e.g., sparse autoencoders, can be scaled to produce many steering vectors, but…
Changing the behavior of large language models (LLMs) can be as straightforward as editing the Transformer's residual streams using appropriately constructed "steering vectors." These modifications to internal neural activations, a form of…
Sparse activation, which selectively activates only an input-dependent set of neurons in inference, is a useful technique to reduce the computing cost of Large Language Models (LLMs) without retraining or adaptation efforts. However,…
The rise of large language models (LLMs) has prompted increasing interest in their use as in-context learning agents. At the core of agentic behavior is the capacity for exploration, or the ability to actively gather information about the…
Autonomous agents based on Large Language Models (LLMs) have evolved from reactive assistants to systems capable of planning, executing actions via tools, and iterating over environment observations. However, they remain vulnerable to…
Safety alignment is indispensable for Large Language Models (LLMs) to defend threats from malicious instructions. However, recent researches reveal safety-aligned LLMs prone to reject benign queries due to the exaggerated safety issue,…
Small language models (SLMs) in the 100M-10B parameter range increasingly power production systems, yet whether they possess the internal emotion representations recently discovered in frontier models remains unknown. We present the first…
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…
Fine-tuning large language models (LLMs) to adapt to evolving safety policies is costly and impractical. Mechanistic interpretability enables inference-time control through latent activation steering, yet its potential for precise,…
Event-related potentials (ERPs) extracted from electroencephalography (EEG) data in response to stimuli are widely used in psychological and neuroscience experiments. A major goal is to link ERP characteristic components to subject-level…
Chain-of-Thought (CoT) prompting has improved the reasoning performance of large language models (LLMs), but it remains unclear why it works and whether it is the unique mechanism for triggering reasoning in large language models. In this…
Small large language models (sLLMs) offer the advantage of being lightweight and efficient, which makes them suitable for resource-constrained environments. However, sLLMs often struggle to maintain topic consistency in task-oriented…