Search: “safety”

CAP: Controllable Alignment Prompting for Unlearning in LLMs

arXiv:2604.21251v1 Announce Type: cross Abstract: Large language models (LLMs) trained on unfiltered corpora inherently risk retaining sensitive information, necessitating selective knowledge unlearni

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Local Ai

Conformal Prediction Assessment: A Framework for Conditional Coverage Evaluation and Selection

DGX agent

arXiv:2603.27189v2 Announce Type: replace-cross Abstract: Conformal prediction provides rigorous distribution-free finite-sample guarantees for marginal coverage under the assumption of exchangeabilit

local-aiarxiv-cs-lg

24 Apr 2026

Omission Constraints Decay While Commission Constraints Persist in Long-Context LLM Agents

arXiv:2604.20911v1 Announce Type: cross Abstract: LLM agents deployed in production operate under operator-defined behavioral policies (system-prompt instructions such as prohibitions on credential di

RewardBench 2: Advancing Reward Model Evaluation

arXiv:2506.01937v2 Announce Type: replace Abstract: Reward models are used throughout the post-training of language models to capture nuanced signals from preference data and provide a training target

Serialisation Strategy Matters: How FHIR Data Format Affects LLM Medication Reconciliation

arXiv:2604.21076v1 Announce Type: cross Abstract: Medication reconciliation at clinical handoffs is a high-stakes, error-prone process. Large language models are increasingly proposed to assist with t

A Vision-Language-Action Model for Adaptive Ultrasound-Guided Needle Insertion and Needle Tracking

arXiv:2604.20347v1 Announce Type: cross Abstract: Ultrasound (US)-guided needle insertion is a critical yet challenging procedure due to dynamic imaging conditions and difficulties in needle visualiza

Foundation Models in Biomedical Imaging: Turning Hype into Reality

arXiv:2512.15808v2 Announce Type: replace-cross Abstract: Foundation models (FMs) are driving a prominent shift in biomedical imaging from task-specific models to unified backbone models for diverse t

From Scene to Object: Text-Guided Dual-Gaze Prediction

arXiv:2604.20191v1 Announce Type: cross Abstract: Interpretable driver attention prediction is crucial for human-like autonomous driving. However, existing datasets provide only scene-level global gaz

Large Language Models Outperform Humans in Fraud Detection and Resistance to Motivated Investor Pressure

arXiv:2604.20652v1 Announce Type: new Abstract: Large language models trained on human feedback may suppress fraud warnings when investors arrive already persuaded of a fraudulent opportunity. We test

Open-Architecture End-to-End System for Real-World Autonomous Robot Navigation

arXiv:2410.06239v3 Announce Type: replace Abstract: Enabling robots to autonomously navigate unknown, complex, and dynamic real-world environments presents several challenges, including imperfect perc

local-aiarxiv-cs-ro

OVPD: A Virtual-Physical Fusion Testing Dataset of OnSite Auton-omous Driving Challenge

model-releasesarxiv-cs-ro

arXiv:2604.20423v1 Announce Type: new Abstract: The rapid iteration of autonomous driving algorithms has created a growing demand for high-fidelity, replayable, and diagnosable testing data. However,

QuadPiPS: A Perception-informed Footstep Planner for Quadrupeds With Semantic Affordance Prediction

arXiv:2501.00112v2 Announce Type: replace Abstract: This work proposes QuadPiPS, a perception-informed framework for quadrupedal foothold planning in the perception space. QuadPiPS employs a novel ego

local-aiarxiv-cs-ro

SkillLearnBench: Benchmarking Continual Learning Methods for Agent Skill Generation on Real-World Tasks

arXiv:2604.20087v1 Announce Type: new Abstract: Skills have become the de facto way to enable LLM agents to perform complex real-world tasks with customized instructions, workflows, and tools, but how

WildFireVQA: A Large-Scale Radiometric Thermal VQA Benchmark for Aerial Wildfire Monitoring

arXiv:2604.20190v1 Announce Type: new Abstract: Wildfire monitoring requires timely, actionable situational awareness from airborne platforms, yet existing aerial visual question answering (VQA) bench

3D Foundation Model for Generalizable Disease Detection in Head Computed Tomography

arXiv:2502.02779v3 Announce Type: replace-cross Abstract: Head computed tomography (CT) imaging is a widely-used imaging modality with multitudes of medical indications, particularly in assessing path

Are Large Language Models Economically Viable for Industry Deployment?

arXiv:2604.19342v1 Announce Type: new Abstract: Generative AI-powered by Large Language Models (LLMs)-is increasingly deployed in industry across healthcare decision support, financial analytics, ente

Beyond Explicit Refusals: Soft-Failure Attacks on Retrieval-Augmented Generation

arXiv:2604.18663v1 Announce Type: cross Abstract: Existing jamming attacks on Retrieval-Augmented Generation (RAG) systems typically induce explicit refusals or denial-of-service behaviors, which are

Distillation Traps and Guards: A Calibration Knob for LLM Distillability

arXiv:2604.18963v1 Announce Type: cross Abstract: Knowledge distillation (KD) transfers capabilities from large language models (LLMs) to smaller students, yet it can fail unpredictably and also under

local-aiarxiv-cs-ai

GenerativeMPC: VLM-RAG-guided Whole-Body MPC with Virtual Impedance for Bimanual Mobile Manipulation

model-releasesarxiv-cs-ro

arXiv:2604.19522v1 Announce Type: new Abstract: Bimanual mobile manipulation requires a seamless integration between high-level semantic reasoning and safe, compliant physical interaction - a challeng

Harmful Intent as a Geometrically Recoverable Feature of LLM Residual Streams

arXiv:2604.18901v1 Announce Type: cross Abstract: Harmful intent is geometrically recoverable from large language model residual streams: as a linear direction in most layers, and as angular deviation

Human-Guided Harm Recovery for Computer Use Agents

arXiv:2604.18847v1 Announce Type: new Abstract: As LM agents gain the ability to execute actions on real computer systems, we need ways to not only prevent harmful actions at scale but also effectivel

Hardware

SpikeMLLM: Spike-based Multimodal Large Language Models via Modality-Specific Temporal Scales and Temporal Compression

arXiv:2604.18610v1 Announce Type: cross Abstract: Multimodal Large Language Models (MLLMs) have achieved remarkable progress but incur substantial computational overhead and energy consumption during

hardwarearxiv-cs-ai

Towards Optimal Agentic Architectures for Offensive Security Tasks

arXiv:2604.18718v1 Announce Type: cross Abstract: Agentic security systems increasingly audit live targets with tool-using LLMs, but prior systems fix a single coordination topology, leaving unclear w

Uncertainty Quantification in Detection Transformers: Object-Level Calibration and Image-Level Reliability

arXiv:2412.01782v4 Announce Type: replace-cross Abstract: DETR and its variants have emerged as promising architectures for object detection, offering an end-to-end prediction pipeline. In practice, h

local-aiarxiv-cs-ai

Watch the Weights: Unsupervised monitoring and control of fine-tuned LLMs

arXiv:2508.00161v3 Announce Type: replace-cross Abstract: The releases of powerful open-weight large language models (LLMs) are often not accompanied by access to their full training data. Existing in

AIM 2025 Rip Current Segmentation (RipSeg) Challenge Report

arXiv:2508.13401v3 Announce Type: replace Abstract: This report presents an overview of the AIM 2025 RipSeg Challenge, a competition designed to advance techniques for automatic rip current segmentati

Causally-Constrained Probabilistic Forecasting for Time-Series Anomaly Detection

arXiv:2604.17998v1 Announce Type: new Abstract: Anomaly detection in multivariate time series is a central challenge in industrial monitoring, as failures frequently arise from complex temporal dynami

local-aiarxiv-cs-lg

Cognitive Chain-of-Thought (CoCoT): Structured Multimodal Reasoning about Social Situations

arXiv:2507.20409v2 Announce Type: replace Abstract: Chain-of-Thought (CoT) prompting helps models think step by step. But naive CoT breaks down in visually grounded social tasks, where models must per

From Heads to Neurons: Causal Attribution and Steering in Multi-Task Vision-Language Models

arXiv:2604.17941v1 Announce Type: cross Abstract: Recent work has increasingly explored neuron-level interpretation in vision-language models (VLMs) to identify neurons critical to final predictions.

Fuzzy Encoding-Decoding to Improve Spiking Q-Learning Performance in Autonomous Driving

model-releasesarxiv-cs-lg

arXiv:2604.16436v1 Announce Type: cross Abstract: This paper develops an end-to-end fuzzy encoder-decoder architecture for enhancing vision-based multi-modal deep spiking Q-networks in autonomous driv

Injecting Structured Biomedical Knowledge into Language Models: Continual Pretraining vs. GraphRAG

arXiv:2604.16422v1 Announce Type: new Abstract: The injection of domain-specific knowledge is crucial for adapting language models (LMs) to specialized fields such as biomedicine. While most current a

MedRedFlag: Investigating how LLMs Redirect Misconceptions in Real-World Health Communication

arXiv:2601.09853v2 Announce Type: replace Abstract: Real-world health questions from patients often unintentionally embed false assumptions or premises. In such cases, safe medical communication typic

NTIRE 2026 Rip Current Detection and Segmentation (RipDetSeg) Challenge Report

arXiv:2604.17070v1 Announce Type: new Abstract: This report presents the NTIRE 2026 Rip Current Detection and Segmentation (RipDetSeg) Challenge, which targets automatic rip current understanding in i

On Inverse Problems, Parameter Estimation, and Domain Generalization

model-releasesarxiv-cs-lg

arXiv:2506.06024v2 Announce Type: replace-cross Abstract: Signal restoration and inverse problems are key elements in most real-world data science applications. In the past decades, with the emergence

RoTRAG: Rule of Thumb Reasoning for Conversation Harm Detection with Retrieval-Augmented Generation

arXiv:2604.17301v1 Announce Type: new Abstract: Detecting harmful content in multi turn dialogue requires reasoning over the full conversational context rather than isolated utterances. However, most

Rule-VLN: Bridging Perception and Compliance via Semantic Reasoning and Geometric Rectification

arXiv:2604.16993v1 Announce Type: cross Abstract: As embodied AI transitions to real-world deployment, the success of the Vision-and-Language Navigation (VLN) task tends to evolve from mere reachabili

TowerDataset: A Heterogeneous Benchmark for Transmission Corridor Segmentation with a Global-Local Fusion Framework

arXiv:2604.16848v1 Announce Type: new Abstract: Fine-grained semantic segmentation of transmission-corridor point clouds is fundamental for intelligent power-line inspection. However, current progress

HarmfulSkillBench: How Do Harmful Skills Weaponize Your Agents?

arXiv:2604.15415v1 Announce Type: cross Abstract: Large language models (LLMs) have evolved into autonomous agents that rely on open skill ecosystems (e.g., ClawHub and Skills.Rest), hosting numerous

HiPreNets: High-Precision Neural Networks through Progressive Training

model-releasesarxiv-cs-lg

arXiv:2506.15064v3 Announce Type: replace Abstract: Deep neural networks are powerful tools for solving nonlinear problems in science and engineering, but training highly accurate models becomes chall

LLMs Corrupt Your Documents When You Delegate

arXiv:2604.15597v1 Announce Type: new Abstract: Large Language Models (LLMs) are poised to disrupt knowledge work, with the emergence of delegated work as a new interaction paradigm (e.g., vibe coding

MTR-DuplexBench: Towards a Comprehensive Evaluation of Multi-Round Conversations for Full-Duplex Speech Language Models

arXiv:2511.10262v3 Announce Type: replace-cross Abstract: Full-Duplex Speech Language Models (FD-SLMs) enable real-time, overlapping conversational interactions, offering a more dynamic user experienc

TwoHamsters: Benchmarking Multi-Concept Compositional Unsafety in Text-to-Image Models

arXiv:2604.15967v1 Announce Type: cross Abstract: Despite the remarkable synthesis capabilities of text-to-image (T2I) models, safeguarding them against content violations remains a persistent challen

ECM Contracts: Contract-Aware, Versioned, and Governable Capability Interfaces for Embodied Agents

arXiv:2604.13097v1 Announce Type: cross Abstract: Embodied agents increasingly rely on modular capabilities that can be installed, upgraded, composed, and governed at runtime. Prior work has introduce

17 Apr 2026

Mechanistic Decoding of Cognitive Constructs in LLMs

arXiv:2604.14593v1 Announce Type: new Abstract: While Large Language Models (LLMs) demonstrate increasingly sophisticated affective capabilities, the internal mechanisms by which they process complex

17 Apr 2026

Rethinking AI Hardware: A Three-Layer Cognitive Architecture for Autonomous Agents

arXiv:2604.13757v1 Announce Type: new Abstract: The next generation of autonomous AI systems will be constrained not only by model capability, but by how intelligence is structured across heterogeneou

local-aiarxiv-cs-ai

17 Apr 2026

A Proactive EMR Assistant for Doctor-Patient Dialogue: Streaming ASR, Belief Stabilization, and Preliminary Controlled Evaluation

arXiv:2604.13059v1 Announce Type: new Abstract: Most dialogue-based electronic medical record (EMR) systems still behave as passive pipelines: transcribe speech, extract information, and generate the

16 Apr 2026

Document-tuning for robust alignment to animals

arXiv:2604.13076v1 Announce Type: new Abstract: We investigate the robustness of value alignment via finetuning with synthetic documents, using animal compassion as a value that is both important in i

16 Apr 2026

FieldWorkArena: Agentic AI Benchmark for Real Field Work Tasks

arXiv:2505.19662v3 Announce Type: replace-cross Abstract: This paper introduces FieldWorkArena, a benchmark for agentic AI targeting real-world field work. With the recent increase in demand for agent