Search: “safety”

Differential Privacy Protected Preference Alignment on Survey Derived Data

arXiv:2610.16458v1 Announce Type: cross Abstract: Preference alignment enables language models to better reflect human values, but the preference data contains sensitive information, especially when d

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

DiffNash: Finding Nash Equilibrium in High-Dimensional Continuous Action Spaces with Diffusion Policies

DGX agent

arXiv:2610.15932v1 Announce Type: new Abstract: Computing equilibria in high-dimensional continuous games requires learning and representing rich, multi-modal strategy distributions, something convent

safetyarxiv-cs-lg

14 Oct 2026

Diffusion-Based Auctions for Multi-Winner Banner Ads

arXiv:2610.14473v1 Announce Type: cross Abstract: We propose a new framework for banner ad auctions, leveraging diffusion models to allow for multiple bidders to co-win a single banner ad opportunity,

Diffusion-Based Posterior Sampling: A Feynman-Kac Analysis of Bias and Stability

arXiv:2605.06538v2 Announce Type: replace Abstract: Diffusion-based posterior samplers draw from measurement- or reward-conditioned posteriors using pretrained diffusion priors. Though widely used for

Disentangling Task-Relevant Latent Dynamics for Reward Modeling

arXiv:2610.14159v1 Announce Type: cross Abstract: Reward design remains a major bottleneck for reinforcement learning in robotic manipulation. Handcrafted dense rewards can compensate for the limited

Dissecting the Carbon-SLO Tension in Geo-Distributed LLM Serving

arXiv:2610.14571v1 Announce Type: cross Abstract: Large Language Model (LLM) inference is driving a growing share of datacenter carbon emissions. Although carbon-aware scheduling can reduce these emis

Distance Marching for Generative Modeling

arXiv:2602.02928v2 Announce Type: replace Abstract: Modern generative models based on diffusion or flow matching transport noise toward the data manifold, without representing the data manifold explic

Diversity You Can Actually Measure: A Fast, Model-Free Diversity Metric for Robotics Datasets

arXiv:2603.11634v2 Announce Type: replace Abstract: Robotics datasets for imitation learning typically consist of long-horizon trajectories of different lengths over states, actions, and high-dimensio

Do My Samples Help? Compute Budget Allocation for Real-World Sampling-Based MPC in Contact-Rich Manipulation

arXiv:2610.14064v1 Announce Type: new Abstract: Sampling-based model predictive control (SMPC) is increasingly practical for contact-rich manipulation because GPU-based physics simulators enable real-

Does self-distillation improve reasoning?

arXiv:2610.18613v1 Announce Type: new Abstract: Self-distillation is an emerging approach that trains a model to imitate its own predictions when conditioned on privileged information, such as correct

DPOison-VL: Clean Preference Poisoning in Vision-Language Direct Preference Optimization

arXiv:2610.18609v1 Announce Type: cross Abstract: Direct Preference Optimization (DPO) has become a widely adopted approach for aligning vision-language models (VLMs), yet the security of multimodal p

DR-APO: Decoupled-Representation Advantage-Guided Policy Optimization from Continuous Control to Embodied Navigation

arXiv:2610.16618v1 Announce Type: new Abstract: Policy representations in reinforcement learning have expanded beyond Gaussian policies to diffusion and flow-matching based generative policies. Howeve

DreamWorld: Unified World Modeling in Video Generation

arXiv:2603.00466v2 Announce Type: replace Abstract: Despite impressive progress in video generation, existing models remain limited to surface-level plausibility, lacking a coherent and unified unders

Drift, Then Spread: Learning One-Step Generative Policies for Online Reinforcement Learning

arXiv:2610.15662v1 Announce Type: new Abstract: Generative policies offer expressive action distributions for online reinforcement learning (RL), but many diffusion- and flow-based policies require co

Drive3R: Block-wise Streaming Geometric Foundation Models for Autonomous Driving

arXiv:2610.15112v1 Announce Type: new Abstract: High-fidelity closed-loop simulation for autonomous driving increasingly relies on dense multi-view 3D reconstruction. While recent feed-forward geometr

EarthAlign: Learning Language-Aligned Representations of Earth System Data

arXiv:2610.14624v1 Announce Type: new Abstract: Interpreting Earth system conditions is central to understanding environmental risks and informing decisions across weather- and climate-sensitive secto

EAU: Environment-Agent-User Co-evolution for Training User-Centric Agents

arXiv:2610.16523v1 Announce Type: cross Abstract: Training LLM agents to interact with users over multiple turns is challenging because the learning signal depends not only on the agent, but also on t

Efficient Frechet Distributional Decoder Alignment for Latent Generative Models

arXiv:2610.17466v1 Announce Type: new Abstract: Frechet Distributional (FD) objectives provide an effective distribution-level signal for post-training generative models. For latent generative models,

EgoAlign: Situated 3D Understanding without Situation Descriptions

arXiv:2610.16176v1 Announce Type: new Abstract: Situated 3D Question Answering requires models to interpret a 3D scene relative to an agent's pose. Existing methods commonly convey that pose through a

EgoMatrix: Physically Faithful Dexterous Robot Demonstrations from Egocentric Human Videos

arXiv:2610.18822v1 Announce Type: new Abstract: Egocentric human videos offer a scalable alternative to robot teleoperation for collecting dexterous manipulation data. However, two challenges limit th

EmbodiedEvo: Self-improving Coding Agents for Physical Intelligence

arXiv:2610.18867v1 Announce Type: new Abstract: Bringing general-purpose agents into the physical world calls for an executable representation of physical intelligence. Compared with vision-language-a

EmbodiSteer: Steering Embodiment-Agnostic Visuomotor Policies with Joint-Space Guidance for Zero-Shot Cross-Embodiment Deployment

arXiv:2606.12965v2 Announce Type: replace Abstract: Scalable robot imitation learning relies on large-scale heterogeneous data from diverse robots or body-free data, making Cartesian end-effector acti

Enabling Token-wise Importance Sampling in Off-Policy RLVR with Proximal KL Regularization

arXiv:2610.18738v1 Announce Type: new Abstract: Off-policy data in reinforcement learning with verifiable rewards (RLVR) can arise from asynchronous generation, rollout-training mismatch, and response

End-to-End Spoken Grammatical Error Correction

arXiv:2506.18532v2 Announce Type: replace-cross Abstract: Grammatical Error Correction (GEC) and feedback play a vital role in supporting second language (L2) learners, educators, and examiners. While

Ensemble-MIX: Enhancing Sample-Efficiency in Multi-Agent RL via Selective Ensemble-Based Exploration

arXiv:2506.02841v5 Announce Type: replace-cross Abstract: Multi-agent reinforcement learning (MARL) methods have achieved state-of-the-art results on a range of multi-agent tasks. Yet, MARL algorithms

Entropy-Regularized Mean-Field Games with Average-Reward Criterion: Span-Seminorm Contractions and Model-Free Learning

arXiv:2610.15871v1 Announce Type: cross Abstract: We study stationary discrete-time mean-field games with finite state and action spaces under the average-reward criterion, with an entropy-regularized

EqC: Equilibrium Energy Control for Introspective Lifelong Robot Learning

arXiv:2610.18790v1 Announce Type: new Abstract: Long-term robot autonomy depends on more than effective action generation: a policy must recognize when its behavior is no longer sufficiently supported

Equilibrium contrastive learning for imbalanced image classification

arXiv:2602.09506v4 Announce Type: replace Abstract: Contrastive learning (CL) is a predominant technique in image classification, but they showed limited performance with an imbalanced dataset. Recent

Esports World Model Dataset

arXiv:2610.16166v1 Announce Type: new Abstract: Interactive world models must capture how multiple participants jointly change a shared environment while observing different parts of it. Competitive g

Estimating the Causal Impact of Social Navigation Signals in Corridor Scenarios

arXiv:2610.16411v1 Announce Type: new Abstract: Deploying robots in human-shared environments requires deciding not only how to navigate, but also when dedicated social signals are worth emitting. Cla

EventLens: Event-Structure Reinforcement Learning for Video Understanding

arXiv:2610.16172v1 Announce Type: new Abstract: Recent multimodal large language models (MLLMs) achieve strong results on video question answering, yet often fail to recover the temporal structure tha

Explore with Memory, Learn without It: In-Context Exploration and Self-Distillation for LLM Reasoning

arXiv:2610.17734v1 Announce Type: new Abstract: Exploration bounds learning efficiency in reinforcement learning with verifiable rewards (RLVR). Current methods explore mostly through independent samp

Fairness under Prediction Error in Predict-then-Optimize Pipelines

arXiv:2610.18593v1 Announce Type: new Abstract: Fairness constraints in predict-then-optimize pipelines are necessarily enforced on predicted outcomes hat{y}, since true outcomes y are unobserved at d

Feature-Context Consistency: Unsupervised Adversarial Detection with Drift Stability on Attributed Graphs

arXiv:2610.16694v1 Announce Type: new Abstract: Adversarial detection on evolving attributed graphs faces two challenges: fraud labels arrive late or not at all, and natural distributional drift erode

FENAS: A Few-Shot Meta-Ensembling Framework for Efficient Neural Architecture Search with Synergized Zero-Cost Proxies

arXiv:2610.15307v1 Announce Type: new Abstract: Efficient neural architecture search (NAS) requires accurate performance estimation under limited evaluation budgets. Existing zero-cost proxies (ZCPs)

Fitted Q-Iteration with Continuous Actions and Adaptive Data: Smoothed Coverage and Minimax Rates

arXiv:2605.05791v2 Announce Type: replace Abstract: Fitted Q-iteration (FQI) is well understood when the data come from a fixed distribution and the action set is finite, whereas value-based continuou

Flow Policy Gradients for Robot Control

arXiv:2602.02481v2 Announce Type: replace Abstract: Likelihood-based policy gradient methods are the dominant approach for training robot control policies from rewards. These methods rely on different

Flow-Support-Weighted Proposals for Test-Time Verified VLA Policies

arXiv:2610.14588v1 Announce Type: cross Abstract: Test-time sampling promises to turn one action prediction into a set of distinct executable candidates that a verifier can choose among. Downstream fi

Force-Modulated Dexterity: Unified Compliance for Contact-Rich Manipulation

arXiv:2610.17161v1 Announce Type: new Abstract: Teleoperation is widely used to collect high-quality demonstrations for robot learning. Dexterous manipulation, however, requires coordinated control of

ForeTac: Predictive Contact Guidance for Generative Robot Policies

arXiv:2610.18118v1 Announce Type: new Abstract: Visually similar actions can produce contact loss, jamming, slippage, or irreversible damage, while tactile feedback may arrive after a critical transit

Formal-Verification-Guided Bayesian Optimization for Non-Intrusive Last-Mile Coverage Closure

arXiv:2610.17449v1 Announce Type: cross Abstract: Last-mile functional coverage closure for modern chip designs remains costly because a mature UVM environment exposes many controls while direct hits

From Hitter to Player: Separating Motor Execution and Strategy in Competitive Humanoid Table Tennis

arXiv:2610.14067v1 Announce Type: new Abstract: Competitive humanoid table tennis requires two distinct abilities: executing a desired return and deciding where to return the ball. We separate these t

From Orbit Matching to Docking: A Hierarchical Reinforcement-Learning Approach to Multiphase Rendezvous Trajectory Design

arXiv:2610.17468v1 Announce Type: cross Abstract: As the complexity of space missions increases, autonomous spacecraft guidance must accommodate transitions between mission phases with distinct dynami

From Parallel to Sequential: Small-Batch Noise Restores Simplicity Bias of Muon

arXiv:2610.16868v1 Announce Type: new Abstract: Simplicity bias, which can arise from sequential feature learning during training, has been proposed as a general explanation for the success of deep le

From Sound to Sense: An Anatomy of Semantic Supervision for Speech-Driven Visual Grounding in Remote Sensing Imagery

arXiv:2610.16364v1 Announce Type: new Abstract: Speech-driven visual grounding in remote sensing (RSVG) removes the typing bottleneck of text-based interaction and is intended to preserve prosodic int

From Video to Vision: Learning Generalizable Humanoid Carrying from Human Motion

arXiv:2610.14561v1 Announce Type: new Abstract: Most existing approaches rely on video or motion-capture data for reinforcement learning (RL) training; however, all of these methods require an upper-l

From Volpiano to Pixels: 30,000 Lines of OMR Ground Truth for Medieval Chant

arXiv:2610.16862v1 Announce Type: new Abstract: Optical music recognition (OMR) for medieval chant lacks training data linking manuscript images to musical symbols across notational families. Using ex

Frozen Site-Probe Accuracy Is Not a Standalone Harmonization Score: An Audit of Evaluation Practice for Multi-Site MRI

arXiv:2610.14140v1 Announce Type: cross Abstract: Image-level MRI harmonization is often evaluated with a frozen site probe, a site classifier trained on raw scans and applied unchanged to harmonized