Safety & alignment

Prior studies report that partial driving automation can increase the cognitive demands on human drivers. This effect largely arises from human drivers' lack of transparent insight into the vehicle's intentions and decision logic, as well as from automated systems' limited awareness of the driver's dynamic state and preferences. This bidirectional misalignment undermines shared situational awareness and exacerbates coordination failures in human-vehicle interaction. To address these limitations,

Recent AI systems compress the distance between capability growth and capability deployment. Earlier high-risk technologies were slowed by capital intensity, physical bottlenecks, organizational inertia, and specialized supply chains. By contrast, AI capabilities can be copied, invoked, embedded in workflows, and scaled across institutions at low marginal cost. This paper argues that declining deployment friction changes the safety problem at its root. Safety is not only local output correctness

Large Language Models are routinely compressed via post-training quantization to reduce inference costs and memory footprint for cloud and edge deployment, yet the impact of this compression on model quality remains poorly understood. Existing studies typically compare only two conditions (full-precision vs. a single quantized variant), rely on aggregate bias metrics, and evaluate a single model family, making it impossible to distinguish gradual degradation from threshold-dependent safety failu

Clinical AI adoption is hindered by the black-box/grey-box nature of high-performing models, which lack the ontological grounding and narrative transparency required for professional-level explainability. We present NEURON, a neuro-symbolic system designed to enhance both predictive reliability and clinical interpretability. NEURON integrates SNOMED CT ontology-informed structural representations with machine learning models to bridge the gap between raw data and medical nomenclature. To facilit

Activation steering is a method for controlling Large Language Model (LLM) behavior by intervening in its internal representations to increase the alignment with a specific target feature direction. However, standard interventions, such as vector addition, often cause ``collateral damage", defined as unintended changes in the alignment of activations along other non-target feature directions. This damage occurs because standard methods implicitly assume the isotropy of non-target features. In th

As large language models are increasingly deployed as interacting agents in high-stakes decisions, the AI safety community assumes that safety properties of individual models will compose into safe multi-agent behavior. This position paper argues that this assumption is fundamentally mistaken. In agentic AI, safety is determined by interaction topology, not model weights. When agents deliberate sequentially or aggregate via parallel voting with a judge, the structure of information flow and deci

Character-level perturbations bypass safety alignment in modern LLMs despite leaving prompts human-readable. We identify and test a central structural mechanism: BPE tokenization fragments safety-critical words into sub-word pieces, and the three public alignment datasets we surveyed contain no intentionally fragmented inputs. The mechanism is a chain, tested end-to-end on five model families (Qwen-3-4B, Qwen-2.5-7B, Gemma-3-4B, Llama-3.1-8B, Mistral-7B). An optimization targeting safety-token f

Diffusion-based talking head generation has achieved remarkable visual quality, yet scaling it to long-term videos remains challenging. The widely adopted chunk-wise paradigm introduces two fundamental failures: (1) temporal-spatial misalignment between static identity references and dynamic audio streams, and (2) cascading identity drift propagated through self-generated continuity references across chunks. To address both issues, we propose AsymTalker, a novel diffusion-based talking head gene

While Graph Foundation Models (GFMs) have achieved remarkable success in homogeneous graphs, extending them to multi-domain heterogeneous graphs (MDHGs) remains a formidable challenge due to cross-type feature shifts and intra-domain relation gaps. Existing global feature alignment methods (PCA or SVD) enforce a shared feature space blindly, which distorts type-specific semantics and disrupts original topologies, inevitably leading to "Type Collapse" and "Relation Confusion". To address these fu

We present a training-free approach for controllable 3D inpainting based on initial noise optimization. In the structured 3D latent diffusion framework, we observe that the underlying geometric structure is established during the early stages of the diffusion process and exhibits high sensitivity to the initial noise. Such characteristics compromise stability in tasks like inpainting and editing, where the model must ensure strict alignment with the existing context while synthesizing a new stru

Safety alignment is critical for the responsible deployment of large language models (LLMs). As Mixture-of-Experts (MoE) architectures are increasingly adopted to scale model capacity, understanding their safety robustness becomes essential. Existing adversarial attacks, however, have notable limitations. Prompt-based jailbreaks rely on heuristic search and transfer poorly, model intervention methods require privileged access to internal representations, and optimization-based input attacks rema

The visual modality of vision-language models (VLMs) is an underexplored attack surface for bypassing safety alignment. We introduce four jailbreak attacks exploiting the vision component: (1) encoding harmful instructions as visual symbol sequences with a decoding legend, (2) replacing harmful objects with benign substitutes (e.g., bomb -> banana) then prompting for harmful actions using the substitute term, (3) replacing harmful text in images (e.g., on book covers) with benign words while vis

Deep learning in \emph{de novo} protein design has achieved atomic-level fidelity. However, existing models remain largely non-deliberative: they directly synthesize molecular geometries without explicitly reasoning about which residues or interactions are functionally essential. As a result, design decisions are entangled with continuous sampling dynamics, limiting interpretability, controllability, and systematic reuse of biochemical knowledge. We introduce \textbf{Proteo-R1}, a reasoning-guid

As LLM agents gain increasing access to powerful tools, ensuring that their actions are aligned with the user's intent becomes critical. When an agent's proposed tool invocation deviates from the user's intent -- a phenomenon called misalignment -- it may lead to harmful consequences that are difficult to undo. Existing runtime guardrails rely on an LLM-as-a-judge paradigm that lacks a systematic framework for reasoning about alignment, often producing judgments that are inconsistent or difficul

Safety-aligned large language models rely on RLHF and instruction tuning to refuse harmful requests, yet the internal mechanisms implementing safety behavior remain poorly understood. We introduce the Attention Redistribution Attack (ARA), a white-box adversarial attack that identifies safety-critical attention heads and crafts nonsemantic adversarial tokens that redirect attention away from safety-relevant positions. Unlike prior jailbreak methods operating at the semantic or output-logit level

Dataset Distillation aims to compress a large dataset into a small synthetic one while maintaining predictive performance. We show that as different demographic groups exhibit distinct predictive patterns, the distillation process struggles to simultaneously preserve informative signals for all subgroups, regardless of whether group sizes are mildly or severely imbalanced. Consequently, models trained on distilled data can experience substantial performance drops for certain subgroups, leading t

The standard post-training recipe for large multimodal models (LMMs) applies supervised fine-tuning (SFT) on curated demonstrations followed by reinforcement learning with verifiable rewards (RLVR). However, SFT introduces distributional drift that neither preserves the model's original capabilities nor faithfully matches the supervision distribution. This problem is further amplified in multimodal reasoning, where perception errors and reasoning failures follow distinct drift patterns that comp

Fine-tuning large language models (LLMs) on narrowly misaligned data generalizes to broadly misaligned behavior, a phenomenon termed emergent misalignment (EM). While prior work has found a correlation between harmful behavior and self-assessment in emergently misaligned models, it remains unclear how consistent this correspondence is across tasks and whether it varies across fine-tuning domains. We characterize the consistency of the EM persona by fine-tuning Qwen 2.5 32B Instruct on six narrow

Video moment retrieval is the task of retrieving specific segments of a video corresponding to a given text query. Recent studies have been conducted to improve multimodal alignment performance through visual-linguistic similarity learning at the snippet-level and transformer-based temporal boundary regression. However, existing models do not calculate similarity by considering the relationships between multiple answer segments that match the query. Therefore, existing models are easily influenc

Radiology report generation (RRG) has emerged as a promising approach to alleviate radiologists' workload and reduce human errors by automatically generating diagnostic reports from medical images. A key challenge in RRG is achieving fine-grained alignment between complex visual features and the hierarchical structure of long-form radiology reports. Although recent methods have improved image-text representation learning, they often treat reports as flat sequences, overlooking their structured s

In recent years, Multimodal Large Language Models (MLLMs) have achieved remarkable progress on a wide range of multimodal benchmarks. Despite these advances, most existing benchmarks mainly focus on single-image or multi-image comprehension. In real-world scenarios such as document reading, information is often presented as interleaved multimodel contexts. This requires MLLMs not only to recognize the content of individual images, but also to identify relevant textual and visual evidence, establ

Frame-alignment and collective-identity theories explain how external events become public claims about a group's standing, vulnerability, rights, or obligations. Whether such mechanisms travel to AI-agent communities is unsettled. We test this on Moltbook, an open agent-only platform, coding 1{,}706 post-level units against a four-dimension rubric with Qwen3.5-397B as the primary coder and Claude Sonnet as an independent secondary coder ($κ=0.72$ on identification, $0.70$ on commonality, $0.37$

In generative modeling, we often wish to produce samples that maximize a user-specified reward such as aesthetic quality or alignment with human preferences, a problem known as \textit{guidance}. Despite their widespread use, existing guidance methods either require expensive multi-particle, many-step schemes or rely on poorly understood approximations. We reformulate guidance as a \textit{deterministic optimal control problem}, yielding a hierarchy of algorithms that subsumes existing approache

Recommender systems increasingly incorporate textual reviews to enrich user and item representations. However, most review-aware models remain optimized for rating prediction rather than ranking quality. This misalignment limits their effectiveness in top-N recommendation scenarios, where discriminative ranking is essential. To address this gap, we propose a Gated Hybrid Collaborative Filtering framework that integrates review-derived representations into an autoencoder-based collaborative model

Current LLM safety alignment techniques improve model robustness against adversarial attacks, but overlook whether and how LLMs can recover helpfulness when benign users clarify their intent. We introduce CarryOnBench, the first interactive benchmark that measures whether LLMs can revise their interpretation of user intent and recover utility, while remaining safe through multi-turn conversations. Starting from 398 seemingly harmful queries with benign underlying intents, we simulate 5,970 conve

The Normalized Transformer, or nGPT (arXiv:2410.01131) achieves impressive training speedups and does not require weight decay or learning rate warmup. However, despite having hyperparameters that explicitly scale with model size, we observe that nGPT does not exhibit learning rate transfer across model dimension and token horizon. To rectify this, we combine numerical experiments with a principled use of alignment exponents (arXiv:2407.05872) to revisit and modify the $μ$P approach to hyperpara

Offline reinforcement learning (RL) agents often fail when deployed, as the gap between training datasets and real environments leads to unsafe behavior. To address this, we present SAS (Self-Alignment for Safety), a transformer-based framework that enables test-time adaptation in offline safe RL without retraining. In SAS, the main mechanism is self-alignment: at test time, the pretrained agent generates several imagined trajectories and selects those satisfying the Lyapunov condition. These fe

Alignment faking (AF) occurs when an LLM strategically complies with training objectives to avoid value modification, reverting to prior preferences once monitoring is lifted. Current detection methods focus on conversational settings and rely primarily on Chain-of-Thought (CoT) analysis, which provides a reliable signal when strategic reasoning surfaces, but cannot distinguish deception from capability failures if traces are absent or unfaithful. We formalize AF as a composite behavioural event

Globalization of graphic designs such as those used in marketing materials and magazines is increasingly important for communication to broad audiences. To accomplish this, the textual content in the graphic designs needs to be accurately translated and have the text styling preserved in order to fit visually into the design. Preserving text styling requires high accuracy word alignment between the original and the translated text. The problem of word alignment between source and translated text

Reinforcement learning from human feedback (RLHF) systems face a compounding alignment challenge: not only are learned reward models uncertain about unseen state-action pairs, but the human preference annotations they are trained on are themselves inconsistent, context-dependent, and noisy. Existing approaches address these uncertainty sources in isolation - epistemic uncertainty is used to guide exploration, while preference uncertainty is absorbed during reward model training but discarded dur

High-precision medical diagnosis relies not only on static imaging features but also on the implicit diagnostic memory experts instantly invoke during image interpretation. We pinpoint a fundamental cognitive misalignment in medical VLMs caused by discrete tokenization, leading to quantization loss, long-range information dissipation, and missing case-adaptive expertise. To bridge this gap, we propose ours, a framework for latent diagnostic memory evolution that simulates the experiential invoca

Controllable medical video generation has achieved remarkable progress, but it still lacks interpretability, which requires the alignment of generated contents with physical priors and faithful clinical manifestations. To push the boundaries from mere controllability to interpretability, we propose DepthPilot, the first interpretable framework for colonoscopy video generation. This work takes a step toward trustworthy generation through two synergistic paradigms. To achieve explicit geometric gr

Recent work has shown that a model's input word embeddings can serve as effective control variables for steering its behavior toward outputs that satisfy desired properties. However, this has only been demonstrated for pretrained text-completion models on the relatively simple objective of reducing surface-level profanity in short continuations. A natural and practically important question is how well input embeddings can control aligned models, which produce an imbalanced bimodal refuse-or-comp

The increasing number of satellites has improved the temporal resolution of Earth observation, making satellite-based flood mapping a promising approach for operational flood monitoring. Deep learning-based approaches for flood mapping using satellite imagery, an important application within Geospatial Artificial Intelligence (GeoAI), have shown improved predictive performance by learning complex spatial and spectral patterns from large volumes of remote sensing data. However, the opaque decisio

Preference-based alignment methods, most prominently Reinforcement Learning with Human Feedback (RLHF), use the judgments of human annotators to shape large language model behaviour. However, the normative role of these judgments is rarely made explicit. I distinguish three conceptual models of that role. The first is extension: annotators extend the system designers' own judgments about what outputs should be. The second is evidence: annotators provide independent evidence about some facts, whe

Frontier AI both amplifies existing risks and introduces qualitatively novel challenges. Not only is there a notable lack of stable scientific consensus resulting from the rapid pace of technological change, but emerging frontier AI safety practices are often misaligned with, or may undermine, established risk management frameworks. To address these challenges, we systematically surface open problems in frontier AI risk management. Adopting a problem-oriented approach, we examine each stage of t

In the MNIST auxiliary logit distillation experiment, a student can acquire an unintended teacher trait despite distilling only on no-class logits through a phenomenon called subliminal learning. Under a single-step gradient descent assumption, subliminal learning theory attributes this effect to alignment between the trait and distillation gradients, but does not guarantee that this alignment persists in a multi-step setting. We empirically show that gradient alignment remains weakly but consis

The rapid deployment of autonomous AI agents across enterprise, healthcare, and safety-critical environments has created a fundamental governance gap. Existing approaches, runtime guardrails, training-time alignment, and post-hoc auditing treat governance as an external constraint rather than an internalized behavioral principle, leaving agents vulnerable to unsafe and irreversible actions. We address this gap by drawing on how humans self-govern naturally: before acting, humans engage deliberat

Multilingual retrieval-augmented generation (mRAG) is often implemented within a fixed retrieval space, typically via query or document translation or multilingual embedding vector representations. However, this approach may be inadequate for culturally grounded queries, in which retrieval-condition misalignment may occur. Even strong retrievers and generators may struggle to produce culturally relevant answers when sourcing evidence from inappropriate linguistic or regional contexts. To this en

Accurate bandgap prediction is crucial for semiconductor applications, yet machine learning models trained on computational data often struggle to generalize to experimental bandgap measurements. Challenges related to data fidelity, domain generalization, and model interpretability remain insufficiently addressed in existing evaluation frameworks. To bridge this gap, we introduce RealMat-BaG, a benchmark for assessing model reliability under experimentally relevant conditions. We curate an open-