Search: “safety”

AnchorDiff: Topology-Aware Masked Diffusion with Confidence-based Rewriting for Radiology Report Generation

arXiv:2605.17071v1 Announce Type: new Abstract: Radiology report generation (RRG) aims to automatically produce clinically accurate textual reports from medical images. Existing methods predominantly

Safety

Anytime and Difficulty-Adaptive PAC-Bayes for Constrained Density-Ratio Network with Continual Learning Guarantees

DGX agent

arXiv:2605.17212v1 Announce Type: new Abstract: A unified framework for learning under covariate shift is presented, in which a constrained density-ratio network approximates the Radon-Nikodym derivat

safetyarxiv-cs-lg

19 May 2026

Are Multimodal LLMs Ready for Surveillance? A Reality Check on Zero-Shot Anomaly Detection in the Wild

arXiv:2603.04727v2 Announce Type: replace-cross Abstract: Multimodal large language models (MLLMs) have demonstrated impressive general competence in video understanding, yet their reliability for rea

ARROW: Augmented Replay for RObust World models

arXiv:2603.11395v2 Announce Type: replace-cross Abstract: Continual reinforcement learning challenges agents to acquire new skills while retaining previously learned ones with the goal of improving pe

Artificial Intolerance: Stigmatizing Language in Clinical Documentation Skews Large Language Model Decision-Making

arXiv:2605.17228v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly deployed in high-stakes domains such as clinical decision support and medical documentation. However, the

safetyarxiv-cs-cl

As a Soviet historian who has spent years writing about the extreme, repressive control Soviet Communism exercised over its unfortunate citi…

As a Soviet historian who has spent years writing about the extreme, repressive control Soviet Communism exercised over its unfortunate citizens, I find it really hard to bring a similar accusation ag

safetyelon-musk--x

AURORA: Contextual Orthogonalization for Geometric Representation Learning in Healthcare Foundation Models

arXiv:2605.17765v1 Announce Type: new Abstract: Recent healthcare foundation models have achieved strong predictive performance through large scale self supervised learning, yet their latent represent

Automatic Generation of High-Performance RL Environments

arXiv:2603.12145v2 Announce Type: replace-cross Abstract: Translating complex reinforcement learning (RL) environments into high-performance implementations has traditionally required months of specia

AutoRubric-T2I: Robust Rule-Based Reward Model for Text-to-Image Alignment

arXiv:2605.17602v1 Announce Type: new Abstract: Aligning Text-to-Image (T2I) generation models with human preferences increasingly relies on image reward models that score or rank generated images acc

Avoiding Structural Failure Modes in Tabular Fair SSL: Online Primal-Dual Allocation under Confidence Gating

arXiv:2605.16446v1 Announce Type: cross Abstract: Semi-supervised learning (SSL) enables prediction with limited labels, but high-stakes tabular applications (medical, credit, recidivism) require stat

Benchmarking transferability of SSL pretraining to same and different modality segmentation tasks

arXiv:2605.18491v1 Announce Type: new Abstract: Methods: Nine SSL methods spanning four pretext-task families were pretrained from scratch using the same 10{,}412 3D CT scans (1.89~M 2D axial slices)

Beyond Compliance: How AI Could Help Creative Writers by Refusing Them

arXiv:2605.16272v1 Announce Type: cross Abstract: Mainstream creativity support design prioritizes compliant AI for seamless writing interactions, but concerns over inappropriate AI reliance highlight

Beyond Policy Optimization: A Data Curation Flywheel for Sparse-Reward Long-Horizon Planning

arXiv:2508.03018v2 Announce Type: replace Abstract: Large Language Reasoning Models have demonstrated remarkable success on static tasks, yet their application to multi-round agentic planning in inter

Beyond RLHF: A Unified Theoretical Framework of Alignment

arXiv:2506.01523v2 Announce Type: replace Abstract: Alignment via reinforcement learning from human feedback (RLHF) has become the dominant paradigm for controlling the quality of outputs from large l

Beyond Scaling: Agents Are Heading to the Edge

arXiv:2605.18535v1 Announce Type: new Abstract: The bottleneck of useful agentic intelligence has shifted from compressing world knowledge into a single model to executing a coordinated system. This p

Beyond the Final Actor: Modeling the Dual Roles of Creator and Editor for Fine-Grained LLM-Generated Text Detection

arXiv:2604.04932v3 Announce Type: replace Abstract: The misuse of large language models (LLMs) requires precise detection of synthetic text. Existing works mainly follow binary or ternary classificati

safetyarxiv-cs-cl

Beyond Transcripts: Iterative Peer-Editing with Audio Unlocks High-Quality Human Summaries of Conversational Speech

arXiv:2605.17652v1 Announce Type: new Abstract: There are not enough established benchmarks for the task fo speech summarization. Creating new benchmarks demands human annotation, as LLMs could embed

safetyarxiv-cs-cl

BIDO: A Biometric Identity Online Authentication Framework

arXiv:2605.16908v1 Announce Type: cross Abstract: Security systems demand continuous, cryptograph- ically robust identity verification without requiring subjects to carry physical tokens, smart cards,

Body-Grounded Perspective Formation and Conative Attunement in Artificial Agents

arXiv:2605.16728v1 Announce Type: new Abstract: This paper proposes a minimal architecture for body-grounded perspective formation in artificial agents. Extending prior work, the model introduces an i

Bridging the Intention-Expression Gap: Aligning Multi-Dimensional Preferences via Hierarchical Relevance Feedback in Text-to-Image Diffusion

arXiv:2603.14936v3 Announce Type: replace Abstract: Users often possess a clear visual intent but struggle to articulate it precisely in language. This intention-expression gap makes aligning generate

Building Reliable Arithmetic Multipliers Under NBTI Aging and Process Variations

arXiv:2605.18444v1 Announce Type: cross Abstract: Hardware aging poses a significant challenge for integrated circuits (ICs), leading to performance degradation and eventual failure. In this work, we

Canonical Regularisation of Wide Feature-Learning Neural Networks

arXiv:2605.18180v1 Announce Type: cross Abstract: Wide neural networks in the feature-learning regime drive modern deep learning, and yet they remain far less studied than their kernel-regime counterp

CatalyticMLLM: A Graph-Text Multimodal Large Language Model for Catalytic Materials

arXiv:2605.17254v1 Announce Type: new Abstract: Property prediction and inverse structural design of catalytic materials are typically modeled as two independent tasks: the former predicts target prop

ChartDesign: Towards LLM Designer of Data Visualization

arXiv:2605.16274v1 Announce Type: cross Abstract: Charts are the dominant medium for visualizing data, discovering patterns and trends, and communicating data driven insights, yet designing them still

ChemVA: Advancing Large Language Models on Chemical Reaction Diagrams Understanding

arXiv:2605.17214v1 Announce Type: new Abstract: While Large Language Models (LLMs) have revolutionized scientific text processing, they exhibit a significant capability gap when interpreting chemical

ClaHF: A Human Feedback-inspired Reinforcement Learning Framework for Improving Classification Tasks

arXiv:2605.17458v1 Announce Type: new Abstract: Text classification models are typically trained via supervised fine-tuning (SFT). However, SFT essentially performs behavior cloning from instance-wise

CodeBind: Decoupled Representation Learning for Multimodal Alignment with Unified Compositional Codebook

arXiv:2605.18257v1 Announce Type: cross Abstract: Multimodal representation alignment is pivotal for large language models and robotics. Traditional methods are often hindered by cross-modal informati

Collaborative Learning for Semi-Supervised LiDAR Semantic Segmentation

arXiv:2605.17135v1 Announce Type: new Abstract: Annotating large-scale LiDAR point clouds for 3D semantic segmentation is costly and time-consuming, which motivates the use of semi-supervised learning

COLSON: Controllable Learning-Based Social Navigation via Diffusion-Based Reinforcement Learning

arXiv:2503.13934v2 Announce Type: replace-cross Abstract: Mobile robot navigation in dynamic environments with pedestrian traffic is a key challenge in the development of autonomous mobile service rob

Confidence-Gated Robot Autonomy: When Does Uncertainty Actually Help?

arXiv:2605.18045v1 Announce Type: cross Abstract: Robotic systems often use predictive uncertainty to decide whether to act autonomously or defer to a fallback policy. In threshold-gated autonomy, unc

Consent Chain Degradation in Embodied Multi-Agent Systems: Bridging the Gap Between AI Agent Governance and Robot Ethics

arXiv:2605.16300v1 Announce Type: cross Abstract: Robotic systems are moving from isolated platforms to interconnected multi-agent ecosystems that operate in human environments. This shift raises a go

Contrastive Conceptor Activation Steering (COAST): Unlocking Vision-Language-Action Models through Hidden States

arXiv:2605.17144v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models leverage powerful perceptual priors from web-scale Vision-Language Model (VLM) pre-training, yet they remain surpr

Controlling Decision Drift in Multimodal Sentiment Analysis with Missing Modalities

arXiv:2605.16889v1 Announce Type: new Abstract: Multimodal sentiment analysis relies on textual, acoustic, and visual signals, yet real-world data often suffer from modality missing and quality imbala

Convex Dataset Valuation for Post-Training

arXiv:2605.16704v1 Announce Type: new Abstract: Improving LLM performance on downstream tasks sometimes requires leveraging auxiliary datasets during post-training. In practice, however, developers fa

COOPO: Cyclic Offline-Online Policy Optimization Algorithm

arXiv:2605.18675v1 Announce Type: cross Abstract: Offline reinforcement learning struggles with distributional shift and constrained performance due to static dataset limitations, while online RL dema

Cracks in the Foundation: A Civil Infrastructure Dataset to Challenge Vision Foundation Models

arXiv:2605.18413v1 Announce Type: new Abstract: Automated structural health monitoring is essential to prevent catastrophic infrastructure failures. Precise, pixel-level defect segmentation is needed

Cross-modal Affinity-aligned Multimodal Learning Analytics for Predicting Student Collaboration Satisfaction in Game-Based Learning

arXiv:2605.16806v1 Announce Type: cross Abstract: Collaborative game-based learning environments offer rich opportunities for small-group knowledge construction, yet automatically predicting student c

Curriculum Group Policy Optimization: Adaptive Sampling for Unleashing the Potential of Text-to-Image Generation

arXiv:2605.17807v1 Announce Type: cross Abstract: Text-to-Image (T2I) generation has achieved remarkable progress in recent years. Meanwhile, reinforcement learning methods, particularly those based o

DACA-GRPO: Denoising-Aware Credit Assignment for Reinforcement Learning in Diffusion Language Models

arXiv:2605.16342v1 Announce Type: cross Abstract: Diffusion large language models are a compelling alternative to autoregressive models, yet existing RL methods for diffusion treat all denoising steps

Dance Across Shifts: Forward-Facilitation Continual Test-Time Adaptation through Dynamic Style Bridging

arXiv:2605.18608v1 Announce Type: new Abstract: Continual Test-Time Adaptation (CTTA) aims to empower perception systems to handle dynamic distribution shifts encountered after deployment. Existing me

DARC: Disagreement-Aware Alignment via Risk-Constrained Decoding

arXiv:2603.08145v2 Announce Type: replace-cross Abstract: Preference-based alignment methods (e.g., RLHF, DPO) typically optimize a single scalar objective, implicitly averaging over heterogeneous hum

Decision-Aware Proximal Bridge Learning for Optimal Treatment Selection

arXiv:2605.16989v1 Announce Type: new Abstract: Individualized treatment selection with continuous actions requires accurate causal response estimation in decision-relevant regions, rather than unifor

Decoupling KL and Trajectories: A Unified Perspective for SFT, DAgger, Offline RL, and OPD in LLM Distillation

arXiv:2605.16826v1 Announce Type: cross Abstract: Knowledge distillation is central to LLM post-training, yet its design space remains poorly understood, especially alongside reinforcement learning (R

Deep Reinforcement Learning Framework for Diversified Portfolio Management Across Global Equity Markets

arXiv:2605.17307v1 Announce Type: cross Abstract: This study develops and evaluates a deep reinforcement learning framework for dynamic portfolio allocation across global equity markets. The Soft Acto

Deep sequence models tend to memorize geometrically; it is unclear why

arXiv:2510.26745v3 Announce Type: replace-cross Abstract: Deep sequence models are said to store atomic facts predominantly in the form of associative memory: a brute-force lookup of co-occurring enti

Model Releases

DeTrack: A Benchmark and Altitude-Aware Dual World Model for Drone-embodied Tracking

model-releasesarxiv-cs-cv

arXiv:2605.17451v1 Announce Type: new Abstract: Aerial object tracking has broad applications in public safety, emergency rescue, wildlife monitoring, and related fields. However, existing aerial trac

DEVIS-GRPO: Unleashing GRPO on Dynamic Extreme View Synthesis

arXiv:2605.16937v1 Announce Type: new Abstract: Trajectory-controlled video generation has become essential for controllable video generation. While current methods perform well under small-view camer

Dexora: Open-source VLA for High-DoF Bimanual Dexterity

arXiv:2605.18722v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models have recently become a central direction in embodied AI, but current systems are restricted to either dual-gripper c

safetyarxiv-cs-ro