Related papers: Breaking Bad Tokens: Detoxification of LLMs Using …

In this paper, we introduce a novel technique for content safety and prompt injection classification for Large Language Models. Our technique, Layer Enhanced Classification (LEC), trains a Penalized Logistic Regression (PLR) classifier on…

We introduce aligned probing, a novel interpretability framework that aligns the behavior of language models (LMs), based on their outputs, and their internal representations (internals). Using this framework, we examine over 20 OLMo,…

Large language models (LLMs) can elicit social bias during generations, especially when inference with toxic prompts. Controlling the sensitive attributes in generation encounters challenges in data distribution, generalizability, and…

As large language models (LLMs) become more integrated into societal systems, the risk of them perpetuating and amplifying harmful biases becomes a critical safety concern. Traditional methods for mitigating bias often rely on data…

Existing detoxification methods for large language models mainly focus on post-training stage or inference time, while few tackle the source of toxicity, namely, the dataset itself. Such training-based or controllable decoding approaches…

Large language models (LLMs) increasingly operate on long inputs, yet their behavior when harmful sentences are sparsely embedded within such inputs remains poorly understood. We present a sensitivity analysis that probes how LLMs extract…

The common toxicity and societal bias in contents generated by large language models (LLMs) necessitate strategies to reduce harm. Present solutions often demand white-box access to the model or substantial training, which is impractical…

In recent years, the advent of the attention mechanism has significantly advanced the field of natural language processing (NLP), revolutionizing text processing and text generation. This has come about through transformer-based…

Large language models (LLMs) exhibit impressive language capabilities but remain vulnerable to malicious prompts and jailbreaking attacks. Existing knowledge editing methods for LLM detoxification face two major challenges. First, they…

Pretrained neural language models (LMs) are prone to generating racist, sexist, or otherwise toxic language which hinders their safe deployment. We investigate the extent to which pretrained LMs can be prompted to generate toxic language,…

Large Language Models (LLMs) achieve strong linguistic performance, yet their internal mechanisms for producing these predictions remain unclear. We investigate the hypothesis that LLMs encode representations of linguistic constraint…

Despite the state-of-the-art performance of Large Language Models (LLMs), these models often suffer from hallucinations, which can undermine their performance in critical applications. In this work, we propose SAFE, a novel method for…

Sparse autoencoders (SAEs) are commonly used to interpret the internal activations of large language models (LLMs) by mapping them to human-interpretable concept representations. While existing evaluations of SAEs focus on metrics such as…

The safety alignment of Large Language Models (LLMs) is vulnerable to both manual and automated jailbreak attacks, which adversarially trigger LLMs to output harmful content. However, current methods for jailbreaking LLMs, which nest entire…

Despite the remarkable achievements of language models (LMs) across a broad spectrum of tasks, their propensity for generating toxic outputs remains a prevalent concern. Current solutions involving finetuning or auxiliary models usually…

Large language models can resist task-misaligned activation steering during inference, sometimes recovering mid-generation to produce improved responses even when steering remains active. We term this Endogenous Steering Resistance (ESR).…

EEG foundation models achieve state-of-the-art clinical performance, yet the internal computations driving their predictions remain opaque: a barrier to clinical trust. We apply TopK Sparse Autoencoders (SAEs) across three architecturally…

The parallel evolution of Large Language Models (LLMs) with advanced code-understanding capabilities and the increasing sophistication of malware presents a new frontier for cybersecurity research. This paper evaluates the efficacy of…

Sparse autoencoders (SAEs) decompose language model representations into a sparse set of linear latent vectors. Recent works have improved SAEs using language model gradients, but these techniques require many expensive backward passes…

The progress in natural language processing (NLP) using large language models (LLMs) has greatly improved patient information extraction from clinical narratives. However, most methods based on the fine-tuning strategy have limited transfer…