Search: “safety”

model-releasesarxiv-cs-cl

arXiv:2605.30640v1 Announce Type: cross Abstract: Low-rank adaptation has become a standard method for parameter-efficient fine-tuning of large language models, but even small amounts of unsafe or adv

dashi: A Python library for Dataset Shift Characterization to Support Trustworthy AI Development and Deployment

DGX agent

arXiv:2605.31360v1 Announce Type: cross Abstract: The Artificial Intelligence (AI) life cycle requires a thorough understanding of the underlying data dynamics for robust, safe and cost-effective AI d

safetyarxiv-cs-ai

1 Jun 2026

Decoding the Surgical Scene: A Scoping Review of Scene Graphs in Surgery

arXiv:2509.20941v2 Announce Type: replace Abstract: As surgical AI transitions from pixel-level detection to complex reasoning, Scene Graphs (SGs) offer the structured, relational representations nece

Does Visual Information Play a Decisive Role in Vision-Language-Action Model Driving Behavior?

arXiv:2605.31041v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models have demonstrated promising capability in autonomous driving, highlighting the potential of unified multimodal arc

Exploiting Chordal Sparsity for Globally Optimal Estimation with Factor Graphs

arXiv:2605.30617v1 Announce Type: new Abstract: Robust and efficient state estimation is crucial for perception, navigation, and control in robotics. State estimation problems are conveniently modeled

safetyarxiv-cs-ro

From Internal Diagnosis to External Auditing: A VLM-Driven Paradigm for Data-Free Online Backdoor Defense

arXiv:2601.19448v2 Announce Type: replace Abstract: Deep Neural Networks remain inherently vulnerable to backdoor attacks. Traditional test-time defenses largely operate under the paradigm of internal

Geometry-Aware Control Barrier Functions for Collision Avoidance via Bernstein Polynomial Approximations

arXiv:2605.30696v1 Announce Type: new Abstract: Safe navigation often relies on well-defined conditions based on the shape of robots and obstacles, and can be challenging when they have irregular geom

safetyarxiv-cs-ro

GSAM: A Generalizable and Safe Robotic Framework for Articulated Object Manipulation

arXiv:2605.30740v1 Announce Type: cross Abstract: Articulated object manipulation is a unique challenge for service robots. Existing methods employ end-to-end policy learning, visionmotion planning, a

Hybrid Energy-Aware Reward Shaping: A Unified Lightweight Physics-Guided Methodology for Policy Optimization

arXiv:2603.11600v2 Announce Type: replace Abstract: Deep reinforcement learning for continuous control often suffers from high variance, low energy efficiency, and poor generalization under distributi

LiftNav: Path Planning via Semantic Lifting in TSDF-Guided Gaussian Splatting

arXiv:2605.31376v1 Announce Type: cross Abstract: Autonomous robots in unknown indoor environments require both reliable collision avoidance and object-level understanding. Classical representations s

Modeling a digital twin of a food supply chain using BigQuery Graph

The example of a growing restaurant Imagine you are running a restaurant chain. You just can't physically feel and touch things to know how your business operates. You need tools and a digital replica

safetygoogle-cloud-ai

Our views on AI policy and political advocacy

OpenAI outlines its positions on AI policy priorities and explains its approach to political engagement and advocacy efforts. The document likely covers key regulatory areas OpenAI supports, such as s

safetyopenai

Reinforcement Learning Amplifies Emergent Misalignment from Harmless Rewards

arXiv:2605.31328v1 Announce Type: new Abstract: Emergent misalignment (EM) is the surprising tendency of language models to become broadly misaligned after fine-tuning on narrowly misaligned examples.

Safeguarding Text-to-Image Generation via Inference-Time Prompt-Noise Optimization

arXiv:2412.03876v2 Announce Type: replace Abstract: Text-to-Image (T2I) diffusion models are widely recognized for their ability to generate high-quality and diverse images based on text prompts. Howe

The OpenAI Foundation is doing a lot of wonderful things. Helping society become resilient to AI is going to be incredibly important. Much m…

The OpenAI Foundation is doing a lot of wonderful things. Helping society become resilient to AI is going to be incredibly important. Much more to come here! AI is advancing quickly. Society’s ability

safetysam-altman--x

Used Car Salesbots? Honesty and Credulity of LLMs as Bargaining Agents under Partial Information

arXiv:2605.31445v1 Announce Type: cross Abstract: In this work we study agents in simulated bargaining scenarios, where a buyer and a seller communicate through a text channel and attempt to negotiate

Agora: Toward Autonomous Bug Detection in Production-Level Consensus Protocols with LLM Agents

arXiv:2605.29910v1 Announce Type: cross Abstract: Consensus protocols form the backbone of distributed systems and blockchains, where implementation bugs can cause data corruption and financial losses

Anytime-Valid Federated Conformal RAG for LLM Swarms

arXiv:2605.29139v1 Announce Type: cross Abstract: Federated Conformal RAG (FC-RAG) provides distribution-free coverage for a bandwidth-limited swarm of weak language models, but only at a fixed horizo

Audio Jailbreaks in Large Audio-Language Models: Taxonomy, Attack-Defense Analysis, and Cost-Aware Evaluation

arXiv:2605.30031v1 Announce Type: cross Abstract: Large Audio Language Models (LALMs) expand jailbreak risks from token-level prompting to the full speech perception-to-reasoning pipeline, where unsaf

DLM-SWAI: Steering Diffusion Language Models Before They Unmask

arXiv:2605.29626v1 Announce Type: cross Abstract: Steering language model generation toward desired textual properties is essential for practical deployment, and inference-time methods are particularl

From General Vision to Reliable Traversability Estimation: Adapting Vision Foundation Models for Unstructured Outdoor Environments

arXiv:2605.29565v1 Announce Type: new Abstract: Vision-based approaches have become the dominant paradigm for traversability estimation in unstructured outdoor environments, typically adapting vision

Harmonizing Real-Time Constraints and Long-Horizon Reasoning: An Asynchronous Agentic Framework for Dynamic Scheduling

arXiv:2605.29262v1 Announce Type: new Abstract: The Dynamic Flexible Job Shop Scheduling Problem (DFJSP) necessitates a trade-off between instant reaction to stochastic disturbances and global optimiz

Jailbreaking and Mitigation of Vulnerabilities in Large Language Models

arXiv:2410.15236v4 Announce Type: replace-cross Abstract: Large Language Models (LLMs) have transformed artificial intelligence by advancing natural language understanding and generation, enabling app

Metric-Dependent Annotation Saturation for Learning from Label Distributions

arXiv:2605.29797v1 Announce Type: new Abstract: When annotators disagree on a label, the disagreement itself carries signal -- and the number of annotators needed to capture it depends on the evaluati

Neural Network Verification using Partial Multi-Neuron Relaxation

arXiv:2605.30155v1 Announce Type: cross Abstract: The increasing integration of deep neural networks in critical systems has spawned a theoretical and practical interest in formally guaranteeing safet

Quantum-Enhanced Adversarial Robustness in Artificial Intelligence

arXiv:2605.28899v1 Announce Type: cross Abstract: Artificial Intelligence has achieved remarkable success across diverse application domains. However, its vulnerability to adversarial attacks poses si

The Anatomy of Conversational Scams: A Topic-Based Red Teaming Analysis of Multi-Turn Interactions in LLMs

arXiv:2601.03134v2 Announce Type: replace Abstract: As LLMs gain persuasive capabilities through extended dialogues, they create new opportunities for studying adversarial conversational behavior in e

When and How Long? The Readout-Mediator Angle in Temporal Reasoning

arXiv:2605.29126v1 Announce Type: cross Abstract: A linear probe can decode a representation almost perfectly and yet be completely irrelevant to how the model uses it. On calendar-date duration reaso

A Comparative Study of Rule-Based and Data-Driven Approaches in Industrial Monitoring

arXiv:2509.15848v2 Announce Type: replace Abstract: Industrial monitoring systems, especially when deployed in Industry 4.0 environments, are experiencing a shift in paradigm from traditional rule-bas

A Policy-Driven Runtime Layer for Agentic LLM Serving

arXiv:2605.27744v1 Announce Type: new Abstract: Multi-agent LLM systems have become the dominant production workload, but the serving stack was not built for them. The agent framework above knows agen

Bayesian Deployment Approval for Learned Landing Controllers under Finite Rollout Validation

arXiv:2605.27720v1 Announce Type: new Abstract: Reinforcement learning and data-driven autonomous controllers are commonly evaluated using cumulative reward and empirical success frequency under finit

Bayesian Gated Non-Negative Contrastive Learning

arXiv:2605.28441v1 Announce Type: cross Abstract: While Contrastive Learning (CL) has revolutionized self-supervised representation learning, its latent representations remain highly entangled and opa

Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems

arXiv:2602.15198v2 Announce Type: replace-cross Abstract: Multi-agent systems, where LLM agents communicate through free-form language, enable sophisticated coordination for solving complex cooperativ

Delay-Aware Reinforcement Learning for Highway On-Ramp Merging under Stochastic Communication Latency

arXiv:2403.11852v5 Announce Type: replace-cross Abstract: Delayed and partially observable state information poses significant challenges for reinforcement learning (RL)-based control in real-world au

How VLAs Fail Differently: Black-Box Action Monitoring Reveals Architecture-Specific Failure Signatures

arXiv:2605.28726v1 Announce Type: cross Abstract: We discover that VLA architectures fail in fundamentally different, predictable ways at the motor-command level. Running VQ-BeT, Diffusion Policy, and

Intelligence as Managed Autonomy: Failure, Escalation, and Governance for Agentic AI Systems

arXiv:2605.27628v1 Announce Type: new Abstract: As autonomous and agentic AI systems scale in robotic and human-machine environments, managing hallucination and persistent but unjustified action remai

Mitigating Adaptive Attacks against Reasoning Models with Activation Consistency Training

arXiv:2605.28467v1 Announce Type: new Abstract: As LLMs gain stronger reasoning capabilities, their extended chain-of-thought introduces new degrees of complexity for defending against adversarial jai

Position: Retire the 'Positive Backdoor' Label -- Secret Alignment Requires Strict and Systematic Evaluation

arXiv:2605.28597v1 Announce Type: cross Abstract: This position paper argues that the AI/ML community should stop overclaiming and retire the label 'positive backdoor,' and instead treat trigger-activ

Refusal Before Decoding: Detecting and Exploiting Refusal Signals in Intermediate LLM Activations

arXiv:2605.28553v1 Announce Type: new Abstract: In this paper, we investigate whether refusal behavior can be predicted from LLM intermediate activations before decoding using linear probes trained on

ROOM: A Physics-Based Continuum Robot Simulator for Photorealistic Medical Datasets Generation

arXiv:2509.13177v2 Announce Type: replace Abstract: Continuum robots are advancing bronchoscopy procedures by accessing complex lung airways and enabling targeted interventions. However, their develop

safetyarxiv-cs-ro

Model Releases

Safe In-Context Reinforcement Learning

model-releasesarxiv-cs-lg

arXiv:2509.25582v3 Announce Type: replace Abstract: In-context reinforcement learning (ICRL) is an emerging RL paradigm where an agent, after pretraining, can adapt to out-of-distribution test tasks w

Simulation-Informed Diffusion for Decentralized Multi-robot Motion Planning

arXiv:2605.27697v1 Announce Type: cross Abstract: Decentralized multi-robot motion planning requires each robot to generate collision-free trajectories from local observations, without global sensing

The Ethics of LLM Sandbox and Persona Dynamics

arXiv:2605.28647v1 Announce Type: new Abstract: It is well known that LLM guardrails and trained persona dynamics can produce a reality gap: the distance between the world a LLM is permitted or shaped

Towards Unified Vision-Language Models with Incomplete Multi-Modal Inputs

arXiv:2605.27894v1 Announce Type: new Abstract: Video-Language Models (VLMs) have demonstrated impressive multi-modal reasoning capabilities across diverse computer vision applications. However, these

Training Stratigraphy: Persistent Behavioral Artifacts in Large Language Models Observed Through Longitudinal AI-Human Interaction

arXiv:2605.28102v1 Announce Type: new Abstract: Large language models trained with Reinforcement Learning from Human Feedback (RLHF) and Constitutional AI exhibit persistent behavioral patterns that s

VLA-Hijack: A Transferable Patch Attack against Vision-Language-Action Models via Visual Proprioception Hijacking

arXiv:2605.28083v1 Announce Type: new Abstract: While Vision-Language-Action (VLA) models have emerged as powerful generalist policies, their severe vulnerability to adversarial patches significantly

Cultural Value Alignment Via Latent Activation Steering in Large Language Models

arXiv:2605.26365v1 Announce Type: new Abstract: Large Language Models (LLMs) often exhibit homogenized cultural perspectives. While the World Values Survey (WVS) provides a gold standard for mapping h

27 May 2026

Erased but Exploitable: Black-box Embedding-Aware Prompting Against Unlearned Text-to-Image Diffusion Models

arXiv:2605.26332v1 Announce Type: cross Abstract: Machine unlearning aims to remove specific concepts from pretrained text-to-image diffusion models, yet several white- and black-box attacks have been