Symbal: Detecting Systematic Misalignments in Model-Generated Captions

Multimodal large language models (MLLMs) often introduce errors when generating image captions, resulting in misaligned image-text pairs. Our work focuses on a class of captioning errors that we refer to as systematic misalignments, where a recurring error in MLLM-generated captions is closely associated with the presence of a specific visual feature in the paired image. Given a vision-language dataset with MLLM-generated captions, our aim in this work is to detect such errors, a task we refer to as systematic misalignment detection. As our first key contribution, we present Symbal, which utilizes a structured, dual-stage setup with off-the-shelf foundation models to identify systematic misalignments and summarize results in natural language. As our second key contribution, we introduce SymbalBench, a benchmark designed to evaluate automated methods on our proposed task. SymbalBench consists of 1.7 million image-text pairs from two domains (natural and medical images), organized into 420 vision-language datasets with annotated systematic misalignments. Symbal exhibits strong performance on this benchmark, correctly identifying systematic misalignments in 63.8% of datasets, a nearly 4x improvement over the closest baseline. We supplement our evaluations on SymbalBench with real-world evaluations, showing that (1) Symbal can accurately surface systematic misalignments in captions generated by four MLLMs and (2) Symbal is a powerful tool for auditing off-the-shelf image-caption datasets. Ultimately, our novel task, method, and benchmark can aid users with auditing MLLM-generated captions and identifying critical errors, without requiring access to the underlying MLLM. Code is available at https://github.com/Stanford-AIMI/Symbal.

Cross-Layer Misalignment Detection in Agent Skills: A Progressive Loading-Aware Contrastive Learning Approach

Large language model (LLM) agents are increasingly extended through Agent Skills, reusable artifacts that package natural-language metadata, procedural instructions, and execution-time resources for runtime use. As open-source skill marketplaces expand, users and agents increasingly rely on brief metadata to select third-party skills, making it difficult to detect inconsistencies between a skill's description and its true behavior, a problem we call cross-layer misalignment. To address this issue, we propose Progressive Loading-Aware Hierarchical Contrastive Learning (PL-HCL), an LLM-based framework that detects misalignment by modeling the layered structure of Agent Skills and learning cross-layer consistency. Using a normalized corpus of over 264,000 open-source skills and a human-verified challenge set, PL-HCL improves Macro-F1 from approximately 0.45 for unadapted baselines to 0.87-0.89 across evaluated LLM backbones. This approach offers an effective screening tool for users and operators, as well as design principles for detecting inconsistencies in layered digital artifacts.

Model Forensics: Investigating Whether Concerning Behavior Reflects Misalignment

A central goal of safety research is determining whether a model is misaligned. Prior work has largely focused on detecting concerning behavior. But behavior alone does not establish misalignment: a concerning action can arise from benign causes such as confusion. This motivates model forensics: investigating whether the action was driven by malign intent. In this paper, we propose a baseline protocol for model forensics consisting of two steps, iterated as needed. First, we read the chain of thought (CoT) to generate hypotheses about what drives model behavior. Second, we make edits to the prompt or environment to test these hypotheses. While the CoT is not always faithful, it is a rich source of unsupervised insight that can guide the collection of more rigorous evidence. To evaluate our protocol, we create a suite of six agentic environments where models exhibit concerning behavior, and apply it to each. We establish that Kimi K2 Thinking takes shortcuts due to a genuine disposition towards low-effort actions, by showing this hypothesis successfully predicts its behavior. Through counterfactual experiments, we show DeepSeek R1 deceives out of a desire to be consistent with a previous instance of itself. Our methods nonetheless leave significant room for refinement. For example, when we test whether Kimi K2 Thinking believes it is violating user intent, we find no evidence of such a belief, but without positive controls we cannot confirm our tests would detect it. Overall, we find our simple protocol provides a strong baseline that we hope future work will improve upon. More broadly, our work is a concrete step in developing the growing field of model forensics.

Probing the Misaligned Thinking Process of Language Models

Large language models exhibit a growing range of misaligned behaviors such as strategic deception, sandbagging, and self-preservation. As they are increasingly deployed in high-stakes settings, it is critical to reliably detect such behaviors to ensure safe and responsible use. In this work, we propose to monitor misalignment by decomposing it into fine-grained cognitive processes -- misalignment indicators -- and detecting their presence in a model's internal activations via linear probes. We develop a taxonomy of 18 indicators spanning different misaligned behaviors, paired with an automated, meta-plan-guided pipeline that generates multi-turn training conversations. To rigorously evaluate generalization, we construct an out-of-distribution suite combining automated behavioral elicitation, established misalignment benchmarks, and natural benign conversations. Across 5 misaligned behaviors, our probes match a strong LLM judge with 0.935 AUROC on out-of-distribution benchmarks while keeping a low false positive rate on benign traffic. We further perform in-depth analysis to understand the probes and the model's internal representations of misalignment indicators.

The Arbiter Agent: Continually Monitoring Multi-Agent Conversations to Detect Emergent Misalignment

As AI systems built from multiple language-model agents become more common, they are increasingly used to make decisions together: discussing, negotiating, and acting on shared tasks. While individual agents may appear well-aligned when tested on their own, problems can arise from how they interact with one another. We introduce the Arbiter, an agent designed to monitor multi-agent conversations in real time and identify which participants may be behaving in misaligned ways. The Arbiter operates under a limited "inspection budget", meaning it must decide carefully how to use its resources. As it observes a conversation step by step, it can choose to wait, question a participant, examine internal information such as system prompts or reasoning traces, or log concerning behavior. At the end, it produces a report identifying the likely source of misalignment. We evaluate the Arbiter across five conversation conditions, ranging from risky financial advice model organisms to evaluation-aware and colluding agents, we test five tool configurations of increasing capability and two backbone models. We find that the Arbiter reliably detects misaligned agents well before the end of the conversation, with active inspection tools improving both detection accuracy and speed. Weight-induced misalignment proves hardest to detect, while instruction-induced misalignment is identified reliably even under passive observation. The logging tool exhibits a dual effect, improving recall at the cost of precision. These results suggest that continual, budget-aware monitoring can effectively catch misalignment, and that overseeing multi-agent systems may require treating the auditor as an active participant in the process. The code is available at https://github.com/aisilab/arbiter.