相关论文: Detecting and Steering LLMs' Empathy in Action

Alignment in LLMs is more brittle than commonly assumed: misalignment can be triggered by adversarial prompts, benign fine-tuning, emergent misalignment, and goal misgeneralization. Recent evidence suggests that some misalignment behaviors…

Effective safety auditing of large language models (LLMs) demands tools that go beyond black-box probing and systematically uncover vulnerabilities rooted in model internals. We present a comprehensive, interpretability-driven jailbreaking…

Large language model (LLM) agents on multi-step tasks suffer reasoning degradation, looping, drift, stuck states, at rates up to 30% on hard tasks. Current solutions include hard step limits (abrupt) or LLM-as-judge monitoring (10-15%…

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…

Steering vectors work for some concepts and layers but fail for others, and practitioners have no way to predict which setting applies before running an intervention. We introduce the Linear Accessibility Profile (LAP), a per-layer…

Alignment safety research assumes that ethical instructions improve model behavior, but how language models internally process such instructions remains unknown. We conducted over 600 multi-agent simulations across four models (Llama 3.3…

Controlling undesirable Large Language Model (LLM) behaviors, such as the generation of unsafe content or failing to adhere to safety guidelines, often relies on costly fine-tuning. Activation steering provides an alternative for…

Selecting the appropriate model at inference time -- the routing problem -- requires jointly optimizing output quality, cost, latency, and governance constraints. Existing approaches delegate this decision to LLM-based classifiers or…

Activation steering has emerged as a cost-effective paradigm for modifying large language model (LLM) behaviors. Existing methods typically intervene at the block level, steering the bundled activations of selected attention heads,…

Controlling the behavior of large language models (LLMs) at inference time is essential for aligning outputs with human abilities and safety requirements. \emph{Activation steering} provides a lightweight alternative to prompt engineering…

Sophisticated instrumentation for AI systems might have indicators that signal misalignment from human values, not unlike a "check engine" light in cars. One such indicator of misalignment is deceptiveness in generated responses. Future AI…

Large Language Models (LLMs) achieve remarkable performance through pretraining on extensive data. This enables efficient adaptation to diverse downstream tasks. However, the lack of interpretability in their underlying mechanisms limits…

Organizations are increasingly adopting and adapting Large Language Models (LLMs) hosted on public repositories such as HuggingFace. Although these adaptations often improve performance on specialized downstream tasks, recent evidence…

I study whether emotionally framed evaluation follow-ups change both the behavior and the calm-relative internal representations of small, locally deployed language models. Our main benchmark uses Qwen 3.5 0.8B on four impossible-constraint…

Inference-time steering is widely regarded as a lightweight and parameter-free mechanism for controlling large language model (LLM) behavior, and prior work has often suggested that simple activation-level interventions can reliably induce…

Robots operating in shared human environments must not only navigate, interact, and detect their surroundings, they must also interpret and respond to dynamic, and often unpredictable, human behaviours. Although recent advances have shown…

Inference-time LLM alignment methods, particularly activation steering, offer an alternative to fine-tuning by directly modifying activations during generation. Existing methods, however, often rely on non-anticipative interventions that…

The research explores the steerability of Large Language Models (LLMs), particularly OpenAI's ChatGPT iterations. By employing a behavioral psychology framework called OCEAN (Openness, Conscientiousness, Extroversion, Agreeableness,…

Large language models sometimes assert falsehoods despite internally representing the correct answer, failures of honesty rather than accuracy, which undermines auditability and safety. Existing approaches largely optimize factual…

This paper introduces a novel framework for simulating and analyzing how uncooperative behaviors can destabilize or collapse LLM-based multi-agent systems. Our framework includes two key components: (1) a game theory-based taxonomy of…