Related papers: Interpretable Safety Alignment via SAE-Constructed…

Safe exploration is critical for using reinforcement learning (RL) in risk-sensitive environments. Recent work learns risk measures which measure the probability of violating constraints, which can then be used to enable safety. However,…

Large Language Models (LLMs) remain vulnerable to optimization-based jailbreak attacks that exploit internal gradient structure. While Sparse Autoencoders (SAEs) are widely used for interpretability, their robustness implications remain…

We investigate feature universality in Gemma-2 language models (Gemma-2-2B and Gemma-2-9B), asking whether models with a four-fold difference in scale still converge on comparable internal concepts. Using the Sparse Autoencoder (SAE)…

As large language models (LLMs) are overwhelmingly more and more integrated into various applications, ensuring they generate safe responses is a pressing need. Previous studies on alignment have largely focused on general…

Sparse autoencoders (SAEs) are increasingly used to extract activation directions for inference-time steering, but their standard sparsity objective treats latent features as independent. This prior can be poorly matched to high-level…

Lifelong multimodal agents must continuously adapt to new tasks through post-training, but this creates a fundamental tension between acquiring capabilities and preserving safety alignment. We demonstrate that fine-tuning aligned…

Recent breakthroughs in Large Language Models (LLMs) have led to their adoption across a wide range of tasks, ranging from code generation to machine translation and sentiment analysis, etc. Red teaming/Safety alignment efforts show that…

Large language models (LLMs) have achieved remarkable success across many applications, but their ability to generate harmful content raises serious safety concerns. Although safety alignment techniques are often applied during pre-training…

Large language models (LLMs) have been significantly improved by instruction fine-tuning, but still lack transparency and the ability to utilize up-to-date knowledge and information. In this work, we propose search-augmented instruction…

Large Audio Language Models (LALMs) have extended the capabilities of Large Language Models (LLMs) by enabling audio-based human interactions. However, recent research has revealed that LALMs remain vulnerable to harmful queries due to…

Sparse Autoencoders (SAEs) are a prominent tool in mechanistic interpretability (MI) for decomposing neural network activations into interpretable features. However, the aspiration to identify a canonical set of features is challenged by…

Large Language Models (LLMs) are increasingly adopted in high-stakes scenarios, yet their safety mechanisms often remain fragile. Simple jailbreak prompts or even benign fine-tuning can bypass these protocols, underscoring the need to…

Multimodal Large Language Models (MLLMs) pose critical safety challenges, as they are susceptible not only to adversarial attacks such as jailbreaking but also to inadvertently generating harmful content for benign users. While internal…

Safety alignment is a key requirement for building reliable Artificial General Intelligence. Despite significant advances in safety alignment, we observe that minor latent shifts can still trigger unsafe responses in aligned models. We…

Fine-tuning large language models (LLMs) is a common practice to adapt generalist models to specialized domains. However, recent studies show that fine-tuning can erode safety alignment, causing LLMs to respond to harmful or unethical…

AI developers often apply safety alignment procedures to prevent the misuse of their AI systems. For example, before Meta released Llama 2-Chat - a collection of instruction fine-tuned large language models - they invested heavily in safety…

How well are unimodal vision and language models aligned? Although prior work have approached answering this question, their assessment methods do not directly translate to how these models are used in practical vision-language tasks. In…

Sparse Autoencoders (SAEs) have been successfully used to probe Large Language Models (LLMs) and extract interpretable concepts from their internal representations. These concepts are linear combinations of neuron activations that…

Sparse Autoencoders (SAEs) have emerged as a powerful unsupervised method for extracting sparse representations from language models, yet scalable training remains a significant challenge. We introduce a suite of 256 SAEs, trained on each…

Large Language Models (LLMs) have become indispensable in real-world applications. However, their widespread adoption raises significant safety concerns, particularly in responding to socially harmful questions. Despite substantial efforts…