Search: “safety”

LLUMI: Improving LLM Writing Assistance for Mental Health Support with Online Community Feedback

arXiv:2605.30273v1 Announce Type: cross Abstract: Large language models (LLMs) show promise in generating supportive responses for mental health queries, but improving their usefulness, empathy, and s

29 May 2026

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Local Ai

A Paired Testing Protocol for Batch-Conditioned Refusal Robustness in LLM Serving

DGX agent

arXiv:2605.27763v1 Announce Type: new Abstract: Safety evaluations of language models often treat serving configuration as fixed background infrastructure, but batch condition is an untested treatment

local-aiarxiv-cs-lg

28 May 2026

Chance-Constrained MPPI under State and Dynamic Object Prediction Uncertainty and the Evaluation of Collision Risk Calibration

arXiv:2605.28330v1 Announce Type: new Abstract: Chance-constrained Model Predictive Path Integral (MPPI) control is increasingly adopted for navigation in dynamic environments to explicitly bound coll

CRaFT: Circuit-Guided Refusal Feature Selection via Cross-Layer Transcoders

arXiv:2604.01604v2 Announce Type: replace Abstract: While modern LLMs are aligned to refuse harmful requests, it is essential to understand the underlying mechanistic basis of this refusal behavior fo

Grounded Cache Routing for Retrieval-Augmented Generation: When Is It Safe to Reuse an Answer?

arXiv:2605.27494v1 Announce Type: cross Abstract: Modern retrieval-augmented generation(RAG) deployments increasingly rely on caching to reduce token cost and time-to-first-token(TTFT). Prefix-level K

LACUNA: Safe Agents as Recursive Program Holes

arXiv:2605.28617v1 Announce Type: new Abstract: LLM agents increasingly act by writing code, yet a split persists between the runtime that drives the agent and the code the model writes. The runtime o

Model Releases

Models That Know How Evaluations Are Designed Score Safer

model-releasesarxiv-cs-ai

arXiv:2605.28591v1 Announce Type: cross Abstract: The validity of AI safety evaluations depends on models behaving consistently across controlled and deployment settings. Prior work has identified tes

Beyond Fixed Benchmarks and Worst-Case Attacks: Dynamic Boundary Evaluation for Language Models

arXiv:2605.06213v2 Announce Type: replace Abstract: Evaluating large language models (LLMs) today rests on fixed benchmarks that apply the same set of items to any model, producing ceiling and floor e

Bridging Control with Neural Network Verifier alpha-beta-CROWN: A Tutorial

arXiv:2605.26577v1 Announce Type: cross Abstract: Learning-based methods for synthesizing controllers have gained popularity due to their high expressiveness and strong empirical performance. However,

ChainCaps: Composition-Safe Tool-Using Agents via Monotonic Capability Attenuation

arXiv:2605.26542v1 Announce Type: cross Abstract: Tool-using agents increasingly operate in open-ended deployment environments, where they compose file systems, web APIs, code interpreters, and enterp

Modernising Reinforcement Learning-Based Navigation for Embodied Semantic Scene Graph Generation

arXiv:2603.25415v2 Announce Type: replace Abstract: Semantic world models enable embodied agents to reason about objects, relations, and spatial context beyond purely geometric representations. In Org

Provably Safe Motion Planning Under Unknown Disturbances

arXiv:2605.26625v1 Announce Type: new Abstract: We present a provably safe sampling-based motion planning algorithm for robotic systems affected by random disturbances of unknown distribution. We cons

Stochastic Decision Horizons for Constrained Reinforcement Learning

arXiv:2602.04599v2 Announce Type: replace Abstract: We propose stochastic decision horizons (SDH), a theoretically grounded framework for solving constrained RL problems with every-step constraint sat

Auditing medical multi-agent AI reveals risks of false consensus

arXiv:2510.10185v2 Announce Type: replace-cross Abstract: Large language models are increasingly being assembled into medical multi-agent systems that emulate multidisciplinary consultation through sp

Energy Shields for Fairness

arXiv:2605.24926v1 Announce Type: new Abstract: Runtime fairness is not a one-time constraint but a dynamic property evaluated over a sequence of decisions. To ensure fairness at runtime, it is necess

First, do no harm: Breaking suicidogenic echo chambers in media recommendation

arXiv:2605.25258v1 Announce Type: cross Abstract: Recommender systems generally optimises user engagement, but this approach is dangerous in mental health contexts. When vulnerable users show signs of

On the Stability and Realizability of Recurrent Polynomial Surrogate Ternary Logic Gate Networks

arXiv:2605.24649v1 Announce Type: cross Abstract: Recurrent Neural Networks (RNNs) can learn to predict Signal Temporal Logic (STL) verdicts online from partial trajectories, but deploying them as run

Reason--Imagine--Act: Closed-Loop LLM Decision Making with World Models for Autonomous Driving

arXiv:2605.24004v1 Announce Type: new Abstract: Large language models (LLMs) are promising for autonomous driving, but semantics-only decision policies can yield physically unsafe behavior in dynamic

Model Releases

RECTOR: Priority-Aware Rule-Based Reranking for Compliance-Aware Autonomous Driving Trajectory Selection

model-releasesarxiv-cs-ai

arXiv:2605.25095v1 Announce Type: new Abstract: Autonomous driving stacks must pick one trajectory from a multi-modal candidate set; choosing by model confidence ignores safety, traffic-law, and comfo

Referential Security as a New Paradigm for AI Evaluations

arXiv:2605.25673v1 Announce Type: cross Abstract: Security evaluations inherently depend on stable identifiers. Any finding, audit, or regulatory decision must remain attached to the specific artifact

SEIDM: A Safe and Efficient Intelligent Driver Model for Autonomous Driving Behavior

arXiv:2605.23915v1 Announce Type: cross Abstract: The Intelligent Driver Model (IDM) is a cornerstone of Adaptive Cruise Control (ACC), valued for its interpretable parameters and effectiveness in car

RAG-Pull: Turning Retrieval into a Code-Injection Channel via Invisible Unicode Perturbations

arXiv:2510.11195v2 Announce Type: replace-cross Abstract: Retrieval-Augmented Generation (RAG) increases the reliability and trustworthiness of the LLM response and reduces hallucination by eliminatin

25 May 2026

Branch-Stochastic Model Predictive Control for Motion Planning under Multi-Modal Uncertainty with Scenario Clustering

arXiv:2605.22600v1 Announce Type: new Abstract: Motion planning for autonomous driving must account for multi-modal uncertainty in both the intentions and trajectories of surrounding vehicles. Handlin

Learning to Evolve: Multi-modal Interactive Fields for Robust Humanoid Navigation in Dynamic Environments

arXiv:2605.21935v1 Announce Type: new Abstract: Safe manipulation-oriented navigation for humanoid robots requires scene memory that remains reliable under locomotion-induced perceptual distortion, en

Optimus: A Robust Defense Framework for Mitigating Toxicity while Fine-Tuning Conversational AI

arXiv:2507.05660v3 Announce Type: replace-cross Abstract: Customizing Large Language Models (LLMs) on untrusted datasets poses severe risks of injecting toxic behaviors. In this work, we introduce Opt

safetyarxiv-cs-cl

Superhuman Safe and Agile Racing through Multi-Agent Reinforcement Learning

arXiv:2605.22748v1 Announce Type: new Abstract: Autonomous systems have achieved superhuman performance in isolation or simulation, yet they remain brittle in shared, dynamic real-world spaces. This f

Distill to Think, Foresee to Act: Cognitive-Physical Reinforcement Learning for Autonomous Driving

arXiv:2605.21139v1 Announce Type: new Abstract: Current end-to-end autonomous driving models are fundamentally constrained by the behavioral cloning ceiling of imitation learning. While reinforcement

21 May 2026

REFLECTOR: Internalizing Step-wise Reflection against Indirect Jailbreak

arXiv:2605.20654v1 Announce Type: new Abstract: While Large Language Models (LLMs) demonstrate remarkable capabilities, they remain susceptible to sophisticated, multi-step jailbreak attacks that circ

21 May 2026

Neural Configuration-Space Barriers for Manipulation Planning and Control

arXiv:2503.04929v3 Announce Type: replace-cross Abstract: Planning and control for high-dimensional robot manipulators in cluttered dynamic environments require computational efficiency and robust saf

20 May 2026

Surviving the Unseen: Predictive Defense for Novel Multi-Turn Multimodal Attacks

arXiv:2605.18988v1 Announce Type: cross Abstract: The expansion of Multimodal Large Language Models (MLLMs) and their integration into autonomous agentic workflows has introduced a non-stationary atta

20 May 2026

Adversarial Fragility and Language Vulnerability in Clinical AI: A Systematic Audit of Diagnostic Collapse Under Imperceptible Perturbations and Cross-Lingual Drift in Low-Resource Healthcare Settings

arXiv:2605.16993v1 Announce Type: cross Abstract: Current clinical artificial intelligence (AI) systems are evaluated almost exclusively on clean, standardised, English-language inputs, conditions tha

Compress the Context, Keep the Commitments: A Formal Framework for Verifiable LLM Context Compression

arXiv:2605.17304v1 Announce Type: cross Abstract: LLM context is not just tokens; it is a set of commitments. Long-running conversations accumulate goals, constraints, decisions, preferences, tool res

safetyarxiv-cs-cl

Contrastive-SDXL: Annotation-Preserving Night-Time Augmentation for Pedestrian Detection

arXiv:2605.16406v1 Announce Type: new Abstract: Night-time pedestrian detection remains challenging because labelled night-time data are limited and large illumination differences make daytime-only tr

Pedestrian-Aware LLM-Driven Behavioral Planning for Autonomous Vehicles

arXiv:2605.16858v1 Announce Type: cross Abstract: Autonomous Vehicles (AVs) must make reliable decisions in dense urban environments where pedestrian behavior is variable, sometimes abnormal, and ofte

Prediction of Challenging Behaviors Associated with Profound Autism in a Classroom Setting Using Wearable Sensors

arXiv:2605.17618v1 Announce Type: new Abstract: Autism Spectrum Disorder (ASD) is characterized by challenges with social interaction and communication and by restricted or repetitive patterns of thou

SafeDiffusion-R1: Online Reward Steering for Safe Diffusion Post-Training

arXiv:2605.18719v1 Announce Type: new Abstract: Diffusion models have been widely studied for removing unsafe content learned during pre-training. Existing methods require expensive supervised data, e

State Contamination in Memory-Augmented LLM Agents

arXiv:2605.16746v1 Announce Type: new Abstract: LLM agents increasingly rely on persistent state, including transcripts, summaries, retrieved context, and memory buffers, to support long-horizon inter

Local Ai

EVA: Editing for Versatile Alignment against Jailbreaks

arXiv:2605.14750v1 Announce Type: cross Abstract: Large Language Models (LLMs) and Vision Language Models (VLMs) have demonstrated impressive capabilities but remain vulnerable to jailbreaking attacks

local-aiarxiv-cs-ai

15 May 2026

Synthesizing POMDP Policies: Sampling Meets Model-checking via Learning

arXiv:2605.14440v1 Announce Type: new Abstract: Partially Observable Markov Decision Processes (POMDPs) are the standard framework for decision-making under uncertainty. While sampling-based methods s

15 May 2026

AgenticAITA: A Proof-Of-Concept About Deliberative Multi-Agent Reasoning for Autonomous Trading Systems

arXiv:2605.12532v1 Announce Type: cross Abstract: Conventional algorithmic trading systems are grounded in deterministic heuristics or offline-trained statistical models that cannot adapt to the seman

Automated alignment is harder than you think

arXiv:2605.06390v2 Announce Type: replace Abstract: A leading proposal for aligning artificial superintelligence (ASI) is to use AI agents to automate an increasing fraction of alignment research as c

Beyond VMAF: Towards Application-Specific Metrics for Teleoperation Video

arXiv:2605.13525v1 Announce Type: cross Abstract: Automated driving has made remarkable progress, yet situations still arise where human intervention is necessary. Teleoperation provides a scalable so

Certified Robustness under Heterogeneous Perturbations via Hybrid Randomized Smoothing

arXiv:2605.12876v1 Announce Type: new Abstract: Randomized smoothing provides strong, model-agnostic robustness certificates, but existing guarantees are limited to single modalities, treating continu

History Anchors: How Prior Behavior Steers LLM Decisions Toward Unsafe Actions

arXiv:2605.13825v1 Announce Type: new Abstract: Frontier LLMs are increasingly deployed as agents that pick the next action after a long log of prior tool calls produced by the same or a different mod

Learning When to Act: Communication-Efficient Reinforcement Learning via Run-Time Assurance

arXiv:2605.12561v1 Announce Type: new Abstract: Safe reinforcement learning (RL) typically asks extit{what} an agent should do. We ask extit{when} it needs to act, and show that a single policy can jo

Certified Gradient-Based Contact-Rich Manipulation via Smoothing-Error Reachable Tubes

arXiv:2602.09368v2 Announce Type: replace Abstract: Gradient-based methods can efficiently optimize controllers by leveraging differentiable simulation and physical priors. However, contact-rich manip

13 May 2026

Generative AI for Visualizing Highway Construction Hazards Through Synthetic Images and Temporal Sequences

arXiv:2605.11276v1 Announce Type: new Abstract: Highway construction workers face a high risk of serious injury or death. Image-based training materials depicting hazardous scenarios are essential for

13 May 2026

Robust Policy Optimization to Prevent Catastrophic Forgetting

arXiv:2602.08813v2 Announce Type: replace Abstract: Large language models are commonly trained through multi-stage post-training: first via RLHF, then fine-tuned for other downstream objectives. Yet e