Beyond Linear Steering: Unified Multi-Attribute Control for Language Models

Controlling multiple behavioral attributes in large language models (LLMs) at inference time is a challenging problem due to interference between attributes and the limitations of linear steering methods, which assume additive behavior in activation space and require per-attribute tuning. We introduce K-Steering, a unified and flexible approach that trains a single non-linear multi-label classifier on hidden activations and computes intervention directions via gradients at inference time. This avoids linearity assumptions, removes the need for storing and tuning separate attribute vectors, and allows dynamic composition of behaviors without retraining. To evaluate our method, we propose two new benchmarks, ToneBank and DebateMix, targeting compositional behavioral control. Empirical results across 3 model families, validated by both activation-based classifiers and LLM-based judges, demonstrate that K-Steering outperforms strong baselines in accurately steering multiple behaviors.

Precise In-Parameter Concept Erasure in Large Language Models

Large language models (LLMs) often acquire knowledge during pretraining that is undesirable in downstream deployments, e.g., sensitive information or copyrighted content. Existing approaches for removing such knowledge rely on fine-tuning, training low-rank adapters or fact-level editing, but these are either too coarse, too shallow, or ineffective. In this work, we propose PISCES (Precise In-parameter Suppression for Concept EraSure), a novel framework for precisely erasing entire concepts from model parameters by directly editing directions that encode them in parameter space. PISCES uses a disentangler model to decompose MLP vectors into interpretable features, identifies those associated with a target concept using automated interpretability techniques, and removes them from model parameters. Experiments on Gemma 2 and Llama 3.1 over various concepts show that PISCES achieves modest gains in efficacy over leading erasure methods, reducing accuracy on the target concept to as low as 7.7%, while dramatically improving erasure specificity (by up to 31%) and robustness (by up to 38%). Overall, these results demonstrate that feature-based in-parameter editing enables a more precise and reliable approach for removing conceptual knowledge in language models.

SafetyNet: Detecting Harmful Outputs in LLMs by Modeling and Monitoring Deceptive Behaviors

High-risk industries like nuclear and aviation use real-time monitoring to detect dangerous system conditions. Similarly, Large Language Models (LLMs) need monitoring safeguards. We propose a real-time framework to predict harmful AI outputs before they occur by using an unsupervised approach that treats normal behavior as the baseline and harmful outputs as outliers. Our study focuses specifically on backdoor-triggered responses -- where specific input phrases activate hidden vulnerabilities causing the model to generate unsafe content like violence, pornography, or hate speech. We address two key challenges: (1) identifying true causal indicators rather than surface correlations, and (2) preventing advanced models from deception -- deliberately evading monitoring systems. Hence, we approach this problem from an unsupervised lens by drawing parallels to human deception: just as humans exhibit physical indicators while lying, we investigate whether LLMs display distinct internal behavioral signatures when generating harmful content. Our study addresses two critical challenges: 1) designing monitoring systems that capture true causal indicators rather than superficial correlations; and 2)preventing intentional evasion by increasingly capable "Future models''. Our findings show that models can produce harmful content through causal mechanisms and can become deceptive by: (a) alternating between linear and non-linear representations, and (b) modifying feature relationships. To counter this, we developed Safety-Net -- a multi-detector framework that monitors different representation dimensions, successfully detecting harmful behavior even when information is shifted across representational spaces to evade individual monitors. Our evaluation shows 96% accuracy in detecting harmful cases using our unsupervised ensemble approach.

Scaling sparse feature circuit finding for in-context learning

Sparse autoencoders (SAEs) are a popular tool for interpreting large language model activations, but their utility in addressing open questions in interpretability remains unclear. In this work, we demonstrate their effectiveness by using SAEs to deepen our understanding of the mechanism behind in-context learning (ICL). We identify abstract SAE features that (i) encode the model's knowledge of which task to execute and (ii) whose latent vectors causally induce the task zero-shot. This aligns with prior work showing that ICL is mediated by task vectors. We further demonstrate that these task vectors are well approximated by a sparse sum of SAE latents, including these task-execution features. To explore the ICL mechanism, we adapt the sparse feature circuits methodology of Marks et al. (2024) to work for the much larger Gemma-1 2B model, with 30 times as many parameters, and to the more complex task of ICL. Through circuit finding, we discover task-detecting features with corresponding SAE latents that activate earlier in the prompt, that detect when tasks have been performed. They are causally linked with task-execution features through the attention and MLP sublayers.

Same Question, Different Words: A Latent Adversarial Framework for Prompt Robustness

Insensitivity to semantically-preserving variations of prompts (paraphrases) is crucial for reliable behavior and real-world deployment of large language models. However, language models exhibit significant performance degradation when faced with semantically equivalent but differently phrased prompts, and existing solutions either depend on trial-and-error prompt engineering or require computationally expensive inference-time algorithms. In this study, built on the key insight that worst-case prompts exhibit a drift in embedding space, we present Latent Adversarial Paraphrasing (LAP), a dual-loop adversarial framework: the inner loop trains a learnable perturbation to serve as a "latent continuous paraphrase" while preserving semantics through Lagrangian regulation, and the outer loop optimizes the language model parameters on these perturbations. We conduct extensive experiments to demonstrate the effectiveness of LAP across multiple LLM architectures on the RobustAlpaca benchmark with a 0.5%-4% absolution improvement on worst-case win-rate compared with vanilla supervised fine-tuning.

Do Sparse Autoencoders Generalize? A Case Study of Answerability

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 domains, and these features can often manifest differently in each context. We examine this through "answerability"-a model's ability to recognize answerable questions. We extensively evaluate SAE feature generalization across diverse answerability datasets for Gemma 2 SAEs. Our analysis reveals that residual stream probes outperform SAE features within domains, but generalization performance differs sharply. SAE features demonstrate inconsistent transfer ability, and residual stream probes similarly show high variance out of distribution. Overall, this demonstrates the need for quantitative methods to predict feature generalization in SAE-based interpretability.

Trust Me, I'm Wrong: High-Certainty Hallucinations in LLMs

Large Language Models (LLMs) often generate outputs that lack grounding in real-world facts, a phenomenon known as hallucinations. Prior research has associated hallucinations with model uncertainty, leveraging this relationship for hallucination detection and mitigation. In this paper, we challenge the underlying assumption that all hallucinations are associated with uncertainty. Using knowledge detection and uncertainty measurement methods, we demonstrate that models can hallucinate with high certainty even when they have the correct knowledge. We further show that high-certainty hallucinations are consistent across models and datasets, distinctive enough to be singled out, and challenge existing mitigation methods. Our findings reveal an overlooked aspect of hallucinations, emphasizing the need to understand their origins and improve mitigation strategies to enhance LLM safety. The code is available at https://github.com/technion-cs-nlp/Trust_me_Im_wrong .

Rethinking AI Cultural Evaluation

As AI systems become more integrated into society, evaluating their capacity to align with diverse cultural values is crucial for their responsible deployment. Current evaluation methods predominantly rely on multiple-choice question (MCQ) datasets. In this study, we demonstrate that MCQs are insufficient for capturing the complexity of cultural values expressed in open-ended scenarios. Our findings highlight significant discrepancies between MCQ-based assessments and the values conveyed in unconstrained interactions. Based on these findings, we recommend moving beyond MCQs to adopt more open-ended, context-specific assessments that better reflect how AI models engage with cultural values in realistic settings.

Open Problems in Machine Unlearning for AI Safety

As AI systems become more capable, widely deployed, and increasingly autonomous in critical areas such as cybersecurity, biological research, and healthcare, ensuring their safety and alignment with human values is paramount. Machine unlearning -- the ability to selectively forget or suppress specific types of knowledge -- has shown promise for privacy and data removal tasks, which has been the primary focus of existing research. More recently, its potential application to AI safety has gained attention. In this paper, we identify key limitations that prevent unlearning from serving as a comprehensive solution for AI safety, particularly in managing dual-use knowledge in sensitive domains like cybersecurity and chemical, biological, radiological, and nuclear (CBRN) safety. In these contexts, information can be both beneficial and harmful, and models may combine seemingly harmless information for harmful purposes -- unlearning this information could strongly affect beneficial uses. We provide an overview of inherent constraints and open problems, including the broader side effects of unlearning dangerous knowledge, as well as previously unexplored tensions between unlearning and existing safety mechanisms. Finally, we investigate challenges related to evaluation, robustness, and the preservation of safety features during unlearning. By mapping these limitations and open challenges, we aim to guide future research toward realistic applications of unlearning within a broader AI safety framework, acknowledging its limitations and highlighting areas where alternative approaches may be required.

Best-of-N Jailbreaking

We introduce Best-of-N (BoN) Jailbreaking, a simple black-box algorithm that jailbreaks frontier AI systems across modalities. BoN Jailbreaking works by repeatedly sampling variations of a prompt with a combination of augmentations - such as random shuffling or capitalization for textual prompts - until a harmful response is elicited. We find that BoN Jailbreaking achieves high attack success rates (ASRs) on closed-source language models, such as 89% on GPT-4o and 78% on Claude 3.5 Sonnet when sampling 10,000 augmented prompts. Further, it is similarly effective at circumventing state-of-the-art open-source defenses like circuit breakers. BoN also seamlessly extends to other modalities: it jailbreaks vision language models (VLMs) such as GPT-4o and audio language models (ALMs) like Gemini 1.5 Pro, using modality-specific augmentations. BoN reliably improves when we sample more augmented prompts. Across all modalities, ASR, as a function of the number of samples (N), empirically follows power-law-like behavior for many orders of magnitude. BoN Jailbreaking can also be composed with other black-box algorithms for even more effective attacks - combining BoN with an optimized prefix attack achieves up to a 35% increase in ASR. Overall, our work indicates that, despite their capability, language models are sensitive to seemingly innocuous changes to inputs, which attackers can exploit across modalities.