Search: “safety”

Strict Subgoal Execution: Reliable Long-Horizon Planning in Hierarchical Reinforcement Learning

arXiv:2506.21039v3 Announce Type: replace Abstract: Long-horizon goal-conditioned tasks pose fundamental challenges for reinforcement learning (RL), particularly when goals are distant and rewards are

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

Subword boundaries are the second meaningful effect. Adding end-of-subword markers as input embeddings produces a large gain throughout trai…

DGX agent

Subword boundaries are the second meaningful effect. Adding end-of-subword markers as input embeddings produces a large gain throughout training (H3): end-boundaries leak future bytes (whitespace alwa

safetynous-research--x

21 May 2026

SUGAR: A Scalable Human-Video-Driven Generalizable Humanoid Loco-Manipulation Learning Framework

arXiv:2605.20373v1 Announce Type: cross Abstract: Building humanoid robots capable of generalizable whole-body loco-manipulation in the real world remains a fundamental challenge. Existing methods eit

Supervised Latent Restructuring for Small-Data Quantum Learning in Plant Phenomics

arXiv:2605.20413v1 Announce Type: new Abstract: High-dimensional biological data often exhibit a severe mismatch between feature dimensionality and sample size, making reliable classification difficul

SURF: Steering the Scalarization Weight to Uniformly Traverse the Pareto Front

arXiv:2605.20619v1 Announce Type: new Abstract: Scalarization is widely used in multi-objective optimization owing to its simplicity and scalability. In many applications, the goal is to generate solu

SynCB: A Synergy Concept-Based Model with Dynamic Routing Between Concepts and Complementary Neural Branches

arXiv:2605.20908v1 Announce Type: new Abstract: Concept-based (CB) models provide interpretability and support test-time human intervention, while standard neural networks (NN) offer strong task perfo

Synchronization and Turn-Taking in Full-Duplex Speech Dialogue Models

arXiv:2605.20356v1 Announce Type: new Abstract: Full-duplex spoken dialogue models (SDMs) can listen and speak simultaneously, enabling interaction dynamics closer to human conversation than turn-base

safetyarxiv-cs-cl

Task-conditioned probing of instruction-tuned multimodal LLMs: Region-specific brain alignment patterns under naturalistic stimuli

arXiv:2506.08277v3 Announce Type: replace-cross Abstract: Recent voxel-wise multimodal brain encoding studies have shown that multimodal large language models (MLLMs) exhibit a higher degree of brain

safetyarxiv-cs-cl

TelePhysics: Physics-Grounded Multi-Object Scene Generation from a Single Image with Real-Time Interaction

arXiv:2605.20290v1 Announce Type: cross Abstract: Recent generative video models achieve impressive visual quality but remain constrained by limited physical consistency and controllability. Existing

The Economics of AI Inference: Inflation Dynamics, Welfare Costs, and Optimal Monetary Policy under the Inference-Cost Phillips Curve

arXiv:2605.20281v1 Announce Type: cross Abstract: We develop a unified microeconomic and monetary theory of artificial intelligence inference costs and their pass-through to inflation, welfare, and op

The Illusion of Intervention: Your LLM-Simulated Experiment is an Observational Study

arXiv:2605.20767v1 Announce Type: new Abstract: Large language models (LLMs) show potential as simulators of human behavior, offering a scalable way to study responses to interventions. However, becau

safetyarxiv-cs-cl

Time-Prompt: Integrated Heterogeneous Prompts for Unlocking LLMs in Time Series Forecasting

arXiv:2506.17631v4 Announce Type: replace Abstract: Time series forecasting aims to model temporal dependencies among variables for future state inference, holding significant importance and widesprea

Model Releases

TRAM: Test-Time Risk Adaptation with Mixture of Agents

model-releasesarxiv-cs-lg

arXiv:2408.08812v2 Announce Type: replace Abstract: Deployed reinforcement learning agents often face safety requirements that are specified only after training, such as new hazard maps, revised risk

Tutor-Student Reinforcement Learning: A Dynamic Curriculum for Robust Deepfake Detection

arXiv:2603.24139v2 Announce Type: replace Abstract: Standard supervised training for deepfake detection treats all samples with uniform importance, which can be suboptimal for learning robust and gene

Uh-oh, the International Space Station is leaking again

The International Space Station has an ongoing air leak in its Russian PrK module that has been known since 2019 , and this leak has delayed the launch of Axiom Space's private astronaut mission . NAS

safetyars-technica

Uncertainty-Calibrated Explainable Artificial Intelligence for Fetal Ultrasound Plane Classification: A Systematic Review

arXiv:2601.00990v2 Announce Type: replace-cross Abstract: Fetal ultrasound is the cornerstone of antenatal care, and accurate recognition of a small set of standard anatomical planes underpins biometr

Velocityformer: Broken-Symmetry-Matched Equivariant Graph Transformers for Cosmological Velocity Reconstruction

arXiv:2605.21483v1 Announce Type: cross Abstract: Precise measurement of the kinematic Sunyaev-Zel'dovich (kSZ) effect - a probe of the large-scale distribution of baryonic matter, a key observable fo

Verifiable Error Bounds for Physics-Informed Neural Network Solutions of Lyapunov and Hamilton-Jacobi-Bellman Equations

arXiv:2603.19545v2 Announce Type: replace-cross Abstract: Many core problems in nonlinear systems analysis and control can be recast as solving partial differential equations (PDEs) such as Lyapunov a

VLANeXt: Recipes for Building Strong VLA Models

arXiv:2602.18532v2 Announce Type: replace Abstract: Following the rise of large foundation models, Vision-Language-Action models (VLAs) emerged, leveraging strong visual and language understanding fro

Waymo suspends freeway rides and pauses its Atlanta operations, as it updates software to improve performance around construction zones and flooded roadways (Reuters)

Reuters: Waymo suspends freeway rides and pauses its Atlanta operations, as it updates software to improve performance around construction zones and flooded roadways — Alphabet's (GOOGL.O) Waymo said

safetytechmeme

What Semantics Survive the Connector? Diagnosing VLM-to-DiT Alignment in Video Editing

arXiv:2605.20795v1 Announce Type: new Abstract: Flow matching based video generative models have been increasingly relying on prepended Vision-Language Models (VLMs) to handle complex, instruction-bas

When does circular financing stop? Similar to a Ponzi Scheme, when cash outflows become greater than inflows. In this last quarter, $NVDA ca…

When does circular financing stop? Similar to a Ponzi Scheme, when cash outflows become greater than inflows. In this last quarter, NVDA cash grew by only ~600, and now ~95% of its operating cash flow

Why Aggregate Accuracy is Inadequate for Evaluating Fairness in Law Enforcement Facial Recognition Systems

arXiv:2603.28675v2 Announce Type: replace Abstract: Facial recognition systems are increasingly deployed in law enforcement and security contexts, where algorithmic decisions can carry significant soc

Why Ask One When You Can Ask k? Learning-to-Defer to the Top-k Experts

arXiv:2504.12988v5 Announce Type: replace Abstract: Existing Learning-to-Defer (L2D) frameworks are limited to single-expert deferral, forcing each query to rely on only one expert and preventing the

Wonder what fraction of those liking this realized it was a joke 🤣

Wonder what fraction of those liking this realized it was a joke 🤣 There's a lot of excitement about OpenAI's resolution of the unit distance conjecture. In this thread I will discuss the striking app

You can be a cheerleader, or you can be a scientist. Can’t be both. cc: @scaling01

Gary Marcus argues that pursuing cheerleading and science as careers are mutually exclusive paths, suggesting fundamental incompatibility between these pursuits. The post likely critiques the idea of

Yowza! Similar in some ways to the Stanford paper recently on LLMs hallucinating responses to images they never saw.

Yowza! Similar in some ways to the Stanford paper recently on LLMs hallucinating responses to images they never saw. This was an amazing and incredibly damning experiment using Microsoft Copilot, by @

A Geometric Analysis of Sign-Magnitude Asymmetry in a ReLU + RMSNorm Block under Ternary Quantization

arXiv:2605.18933v1 Announce Type: new Abstract: Pre-norm Transformers with RMSNorm tolerate ternary {-1,0,+1} weight quantization with surprisingly small loss (Ma et al., 2024). We give a geometric ex

A Heuristic Approach for Performance Tuning in RL-based Quadrotor Control via Reward Design and Termination Conditions

arXiv:2605.19166v1 Announce Type: cross Abstract: Reinforcement learning (RL)-based quadrotor control policies have achieved impressive performance in tasks such as fast navigation in cluttered enviro

A Unified Framework for Structure-Aware Clustering and Heterogeneous Causal Graph Learning

arXiv:2605.19313v1 Announce Type: cross Abstract: In complex multivariate systems, interactions among variables are defined by dependency structures, often encoded as directed acyclic graphs (ext{DAGs

Accurate Evaluation of Quickest Changepoint Detectors via Non-parametric Survival Analysis

arXiv:2605.18798v1 Announce Type: new Abstract: We propose non-parametric estimators for the average run length (ARL) and average detection delay (ADD) in quickest changepoint detection (QCD) under fi

Aerial Inspection Behaviors via RL-based Quadrotor Control for Under-canopy Forest Environments

arXiv:2605.19202v1 Announce Type: cross Abstract: This paper addresses the problem of using a deep Reinforcement Learning (RL)-based low-level Quadrotor controller within an autonomous Quadrotor navig

Agent Sandbox on GKE is now available for everyone, and a first look at Agent Substrate

In just a short time, we’ve seen AI transition from simple chat interfaces to autonomous agents capable of function calling, code execution, and persistent terminal use. But to orchestrate these capab

safetygoogle-cloud-ai

Atomistic Modeling of Chemical Disorder in Materials: Bridging Classical Methods and AI-Assisted Approaches

arXiv:2605.19124v1 Announce Type: cross Abstract: Chemical disorder, originating from the mixed occupation of crystallographic sites by multiple elements, is widespread in alloys, ceramics, and compos

Automatically Improving Simulation Physics for Articulated Objects

arXiv:2605.19136v1 Announce Type: new Abstract: Simulation is a central tool for scalable robot learning, but its effectiveness depends on the quality of object assets. While modern 3D datasets provid

B-cos GNNs: Faithful Explanations through Dynamic Linearity

arXiv:2605.19778v1 Announce Type: new Abstract: We introduce B-cos GNNs, an inherently explainable class of graph neural networks whose predictions decompose exactly into per-node, per-feature contrib

BERTO: Intent-Driven Network Time Series Forecasting via Natural Language Operator Preferences

arXiv:2512.05721v2 Announce Type: replace Abstract: Traditional cellular traffic forecasting models are optimized for minimizing symmetric errors, leaving them indifferent to shifting operational prio

Beyond Action Residuals: Real-World Robot Policy Steering via Bottleneck Latent Reinforcement Learning

arXiv:2605.19919v1 Announce Type: new Abstract: Pretrained imitation policies have become a strong foundation for robot manipulation, but they often require online improvement to overcome execution er

Beyond Extrapolation: Knowledge Utilization Paradigm with Bidirectional Inspiration for Time Series Forecasting

arXiv:2605.19249v1 Announce Type: new Abstract: Time-series forecasting is critical in various scenarios, such as energy, transportation, and public health. However, most existing forecasters rely pri

Beyond Isotropy in JEPAs: Hamiltonian Geometry and Symplectic Prediction

arXiv:2605.20107v1 Announce Type: cross Abstract: JEPAs often regularize one-view embeddings toward an isotropic Gaussian, implicitly baking Euclidean symmetry into the representation. We show that th

Beyond Mode Collapse: Distribution Matching for Diverse Reasoning

arXiv:2605.19461v1 Announce Type: new Abstract: On-policy reinforcement learning methods like GRPO suffer from mode collapse: they exhibit reduced solution diversity, concentrating probability mass on

Boosting Text-to-Image Diffusion Models via Core Token Attention-Based Seed Selection

arXiv:2605.19532v1 Announce Type: new Abstract: Text-to-image diffusion models can synthesize high-quality images, yet the outcome is notoriously sensitive to the random seed: different initial seeds

Brain alignment of reasoning and action representations from vision-language and action models during naturalistic gameplay

arXiv:2605.19352v1 Announce Type: cross Abstract: Understanding how humans and artificial intelligence systems predict and plan by interacting with their environment is a fundamental challenge at the

Can Large Language Models Revolutionize Survey Research? Experiments with Disaster Preparedness Responses

arXiv:2605.19229v1 Announce Type: new Abstract: Survey research faces mounting structural challenges: declining response rates, sample bias, block-wise missingness among at-risk respondents, and AI-as

CEER: Compliant End-Effector and Root Control as a Unified Interface for Hierarchical Humanoid Loco-Manipulation

arXiv:2605.19981v1 Announce Type: new Abstract: Humanoid robots have achieved impressive locomotion performance, yet contact-rich and long-horizon manipulation remains a major bottleneck. Manipulation

Certifiable Alignment of GNSS and Local Frames via Lagrangian Duality

arXiv:2512.20931v2 Announce Type: replace Abstract: Estimating the absolute orientation of a local system relative to a global navigation satellite system (GNSS) reference often suffers from local min

Chessformer: A Unified Architecture for Chess Modeling

arXiv:2605.19091v1 Announce Type: new Abstract: Chess has long served as a canonical testbed for artificial intelligence, but modeling approaches for its central tasks have diverged. Maximizing playin

CoLD: Counterfactually-Guided Length Debiasing for Process Reward Models in Mathematical Reasoning

arXiv:2507.15698v2 Announce Type: replace-cross Abstract: Process Reward Models (PRMs) play a central role in evaluating and guiding multi-step reasoning in large language models (LLMs), especially fo