Search: “safety”

Towards Bridging the Gap Between Offline and Iterative Alignment via Preference Distillation

arXiv:2609.06893v1 Announce Type: cross Abstract: Direct preference optimization DPO is a promising offline approach for aligning large language models (LLMs) due to its simplicity, computational effi

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

Tracking the Moving Frontier: Long-Short Term Advantage Estimator

DGX agent

arXiv:2609.06671v1 Announce Type: new Abstract: Group-based RLVR methods estimate advantages by repeatedly sampling multiple trajectories for each prompt, making long-horizon agent training expensive

safetyarxiv-cs-lg

9 Sept 2026

TV-Regulated OPD: Direction Matters in On-Policy Distillation

arXiv:2609.08341v1 Announce Type: new Abstract: On-Policy Distillation (OPD) facilitates the transfer of knowledge from domain expert to student in the post-training phase of Large Language Models (LL

Unified Long Video Inpainting and Outpainting via Overlapping High-Order Co-Denoising

arXiv:2511.03272v2 Announce Type: replace Abstract: Diffusion-based text-to-video models are increasingly capable, but mask-based editing over hundreds of frames remains challenging: naive long-video

UniFusion: Sparse-View 4D Reconstruction via Unified Spatio-temporal Depth Alignment

arXiv:2609.05888v1 Announce Type: new Abstract: In this paper, we address the challenging problem of 4D reconstruction from sparse-view videos. This setup usually relies on monocular depth estimation

Vision: Data-Centric Anchoring for Robust and Interpretable Agentic AI

arXiv:2609.08216v1 Announce Type: cross Abstract: Agentic AI systems built on large language models fail in two persistent ways that scaling does not fix: they break under distribution shift, and they

VoT: Vision-of-Thought for Unified Multimodal Representation Alignment

arXiv:2609.07815v1 Announce Type: new Abstract: Current text-to-image systems typically employ a 'text encoder plus diffusion decoder' paradigm, in which text semantics directly modulate continuous la

WeaveLA: Event Driven Cross-Subtask Latent Memory Weaving for Repetitive Robot Manipulation

arXiv:2606.17463v2 Announce Type: replace Abstract: Vision-Language-Action (VLA) policies have achieved remarkable single-step manipulation, yet they remain brittle precisely where each stage depends

When Do Options Help? Policy Necrosis and Redundant Coverage in Option-Critic

arXiv:2609.05508v1 Announce Type: new Abstract: Option-critic learns options: sub-policies together with a learned rule for when each one hands control back. Its headline result is that performance im

When Speech Meets Lips: Interpretable Audio-Visual Synchronization for L2 Pronunciation Assessment

arXiv:2609.06788v1 Announce Type: new Abstract: Automatic Pronunciation Assessment (APA) systems have achieved strong performance with transformer-based models and self-supervised speech representatio

When Superpixels Fail on Documents: A Study of Segmentation for LIME Explanations

arXiv:2609.07462v1 Announce Type: new Abstract: Post-hoc explanation methods are widely used to inspect image classifiers, but their reliability depends on design choices that are often treated as imp

Where Success Breaks: Failure-Boundary Learning for Robust Vision-Language-Action Models

arXiv:2609.06114v1 Announce Type: new Abstract: Vision-language-action (VLA) models adapted through supervised fine-tuning (SFT) inherit a structural asymmetry: expert demonstrations teach the policy

WM-Craftnet: World Synesthesia Model for Generalizable and Robust Dexterous In-Hand Manipulation

arXiv:2609.07002v1 Announce Type: new Abstract: Generalizable and robust dexterous in-hand manipulation requires a policy to infer object pose, geometry, contact, and potential slip from partial and n

Zero-Shot Sim-to-Real Contact-Rich Assembly via Proprioception-Anchored Cross-Modal Pretraining

arXiv:2609.07534v1 Announce Type: new Abstract: Contact-rich assembly remains challenging because it requires submillimeter spatial accuracy and reliable interpretation of forces during sustained cont

ZetaDial: dialing net charge of protein binders at inference time for therapeutic developability

arXiv:2609.05451v1 Announce Type: cross Abstract: Net charge is a developability-relevant property of therapeutic binders, linked to viscosity, clearance, nonspecific interaction and aggregation, and

ZK-Trace: Certified Collusion Tracing with Zero-Knowledge Credentials for Federated GNSS Interference Monitoring

arXiv:2609.08763v1 Announce Type: cross Abstract: Federated global navigation satellite system (GNSS) monitoring distributes a proprietary classifier to partly trusted stations, any of which may leak

A Human-in-the-Loop Framework for AI-Assisted Scoring in Large-Scale Writing Assessment

arXiv:2609.05143v1 Announce Type: cross Abstract: The integration of artificial intelligence (AI), particularly large language models (LLMs), into educational assessment has opened new opportunities t

A Unified Physics-Aware Quantum Machine Learning Framework across Power GaN HEMTs and Logic Nanowire FETs: Predicting Unseen Process Splits and Held-Out Geometry Combinations with Lower Error and Tighter Split-to-Split Variability

arXiv:2609.05251v1 Announce Type: new Abstract: We present a unified reinforcement-learning (RL) framework that discovers compact parametrized quantum circuits (PQCs) for data-scarce device modeling.

A Verifier-Guided Explainable Reasoning Framework with Gold-Anchored QLoRA, Task-Aware Mixture-of-Experts, and Group-Relative RLVR

arXiv:2609.05221v1 Announce Type: cross Abstract: Large language models (LLMs) show strong reasoning ability, but their explanations can remain inconsistent, weakly grounded, or difficult to verify. W

AngelFingerprint: A Traceable, Explainable, and White-Box Stealthy Watermark for Text-Guided Image Editing

arXiv:2609.04709v1 Announce Type: new Abstract: Text-guided diffusion editing raises disinformation concerns, making reliable image provenance essential. While watermarks are commonly used for this pu

BEAM3R: Beam's-eye-view architecture with Mamba-3 for implicit dose reconstruction

arXiv:2609.04747v1 Announce Type: cross Abstract: To enable accurate and rapid photon control point and proton beamlet dose calculation in the DoseRAD2026 challenge, we present BEAM3R, a dose estimati

Beneath the Surface of Chains-of-Thought: A Mechanistic Interpretation of Reasoning Operations in LLMs

arXiv:2609.04753v1 Announce Type: new Abstract: Reasoning in large language models unfolds through diverse functional operations, such as problem formulation, goal decomposition, and deduction. Althou

Commonsense Reasoning in Computer Vision: Foundations, Recent Advancements, and Future Directions

arXiv:2609.05257v1 Announce Type: new Abstract: Commonsense reasoning in computer vision encompasses integrating visual data and contextual knowledge, crucial for enhancing AI's understanding of every

Compositional Reward Models for Conditional Medical Image Generation

arXiv:2609.05028v1 Announce Type: new Abstract: Acquiring high quality annotated medical image data is critical for training deep learning models; however, annotation is expensive, time consuming, and

Consensus Group Relative Policy Optimization for Text Generation

arXiv:2602.03102v2 Announce Type: replace Abstract: Many strong decoding methods for text generation follow a sample-and-rerank paradigm: they draw multiple candidates, score each under a utility (rew

CONTINUITY: Security-Context Contracts for Composable LLM Agent Controls

arXiv:2609.05269v1 Announce Type: cross Abstract: LLM agent systems increasingly combine provenance tracking, authorization, policy enforcement, protocol adapters, and execution controls. However, ind

Controlling and Assessing Appropriate Persona Use in LLM-based Dialogue Generation

arXiv:2609.04676v1 Announce Type: new Abstract: In persona-based dialogue generation (PDG), LLMs often overuse persona attributes by incorporating them regardless of dialogue context, resulting in unn

CoSkill: Joint Reinforcement Learning of Reasoning and Meta-Skill Agents for Hierarchical Skill Evolution

arXiv:2609.04865v1 Announce Type: new Abstract: Skill libraries improve the sample efficiency of agentic reinforcement learning (RL) by enabling large language model (LLM) agents to reuse procedural k

Coupled Control and Wireless World Models for Resilient Remote Robotic Control

arXiv:2609.04851v1 Announce Type: cross Abstract: Remote robotic systems operating over wireless networks must maintain reliable control despite limited communication resources, changing channel condi

Cross-dataset transportability of pediatric chest X-ray deep learning across three countries: discrimination, calibration, operating-point failure, and limited-label recovery

arXiv:2609.05140v1 Announce Type: cross Abstract: Background and Objective: External evaluation of medical-imaging AI is often collapsed into discrimination. We evaluated a computational protocol that

Cultural Misalignment in Large Language Models: Detection, Measurement, and Mitigation Through Targeted Fine-Tuning

arXiv:2609.04485v1 Announce Type: cross Abstract: We evaluate three open-weight LLMs (Gemma3-12B from the USA, Bielik-11B-v3 from Poland, and Qwen3-4B from China) against World Values Survey Wave 7 da

Dressing in Motion: A Human Motion-Aware Diffusion Policy for Robot-Assisted Dressing

arXiv:2609.04759v1 Announce Type: new Abstract: Robotic dressing assistance is a promising solution for supporting older adults with physical impairments in daily living. However, dressing under human

Evidence Integration in Large Language Models

arXiv:2609.04290v1 Announce Type: cross Abstract: Despite increasing reliance on LLMs that reason with external evidence supplied by tools, retrieval-augmented generation, other agents, and users, how

Few-Shot Video Recognition via Hierarchical Metric Learning

arXiv:2609.05242v1 Announce Type: new Abstract: Few-shot action recognition (FSAR) aims to recognize unseen action categories with only a small number of annotated video samples. Recent works typicall

From Matching Models to Recruiting Agents: A Systematized Narrative Review of AI Recruitment Systems, Evaluation, and Governance

arXiv:2609.04286v1 Announce Type: new Abstract: Artificial intelligence in recruitment has shifted the object being automated from profile pairs and ranked lists to multi-stage workflows that retrieve

FSPGD: Rethinking Black-box Attacks on Semantic Segmentation

arXiv:2502.01262v3 Announce Type: replace Abstract: Black-box adversarial attacks on semantic segmentation remain a challenging problem, particularly in the black-box transfer attack setting where per

Gradient-based Model Shortcut Detection for Time Series Classification

arXiv:2510.10075v2 Announce Type: replace-cross Abstract: Deep learning models have attracted lots of research attention in time series classification (TSC) task in the past two decades. Recently, dee

InterSing: Explicit Interaction Dynamics for 3D Duet Singing Animation and Beyond

arXiv:2609.04903v1 Announce Type: new Abstract: We present InterSing, a framework for generating realistic 3D head animations for duet singing performances. Unlike solo singing, duet performance requi

Joint Alignment and Distillation for Video Generation via Sample-Guided Distribution Matching

arXiv:2609.04283v1 Announce Type: new Abstract: Aligning video generative models to human preferences heavily relies on Reinforcement Learning (RL), which suffers from extensive computational overhead

Latent-Aligned Reasoning for Multimodal Recommendation

arXiv:2609.04645v1 Announce Type: cross Abstract: Multimodal Vision-Language Models (VLMs) have demonstrated remarkable capabilities in cross-modal understanding, yet a fundamental challenge persists

Leveraging Imperfect Restoration for Data Availability Attack

arXiv:2609.04627v1 Announce Type: new Abstract: The abundance of online data is at risk of unauthorized usage in training deep learning models. To counter this, various Data Availability Attacks (DAAs

LLM-Assisted Behavioural and Scenario Augmentation for Agent-Based Energy Adoption Models

arXiv:2609.04866v1 Announce Type: new Abstract: Recent advances in large language models (LLMs) create opportunities to enrich simulation-based energy policy analysis, particularly by supporting struc

MCPO: Modality-Contrastive Preference Optimization for Multimodal Chain-of-Thought Compression

arXiv:2609.04947v1 Announce Type: cross Abstract: Recently, multimodal large-scale reasoning models have demonstrated remarkable capabilities in solving complex tasks through long Chains-of-Thought (M

Measuring AI Accountability Through Argumentation Analysis: Can Model Reasoning Withstand Scrutiny?

arXiv:2609.05088v1 Announce Type: new Abstract: AI oversight methods rely on ground truth for validation, but what constitutes appropriate AI behavior is contested. This leaves evaluation of moral rea

Measuring the Novelty of Biomedical Papers Using the Latent Distances between Knowledge Units

arXiv:2609.05175v1 Announce Type: cross Abstract: Measuring the novelty of scientific papers is a central concern in research evaluation and scientometrics. From a recombination perspective, prior stu

MedProb: Probing Internal Representations of Vision-Language Models for Medical Question Answering

arXiv:2609.04336v1 Announce Type: new Abstract: Medical visual question answering (Med-VQA) is often assumed to require medical fine-tuning, large models, or complex multi-agent pipelines. We revisit

MePo++: Unifying Representation Refinement and Reconciliation for General Continual Learning

arXiv:2609.05075v1 Announce Type: new Abstract: General continual learning (GCL) aims to learn from evolving data streams without task identities, explicit boundaries, or repeated access to previous d

MoirfEolas and CriochScore: Developing Resources for and the Evaluation of Tokenization Alignment with Irish Morphology

arXiv:2609.05022v1 Announce Type: new Abstract: This paper presents new tokenization resources for Irish and evaluation measures of alignment with the morphological boundaries of the language. We pres