safety

Large-Language-Models-as-a-Judge in Theory-Agnostic Adaptive Metric-Alignment for Prototypical Networks in Personality Recognition

arXiv:2607.08374v1 Announce Type: cross Abstract: Personality recognition has traditionally been constrained by theory-dependent formulations, where models are trained to fit predefined psychological

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

Latent Memory Palace: Reasoning for Control as Autoregressive Variational Inference

DGX agent

arXiv:2607.08724v1 Announce Type: new Abstract: Human decision-making is highly flexible -- some actions are taken immediately; others require longer deliberation. Language models have exhibited a sim

safetyarxiv-cs-lg

10 Jul 2026

Learning Adaptive Solvers for Distributed Factor Graph Optimization on Matrix Lie Groups

arXiv:2607.08735v1 Announce Type: new Abstract: Modern robotic perception increasingly involves large-scale geometric optimization problems distributed across multiple robots or sessions. However, exi

LESV: Language Embedded Sparse Voxel Fusion for Open-Vocabulary 3D Scene Understanding

arXiv:2604.01388v2 Announce Type: replace Abstract: Recent advancements in open-vocabulary 3D scene understanding heavily rely on 3D Gaussian Splatting (3DGS) to register vision-language features into

LiteOdyssey: A Lightweight Reasoning AI Agent for Interpretable Rare-Disease Diagnosis

arXiv:2606.16149v2 Announce Type: replace Abstract: Rare disease diagnosis involves interpreting clinical and genetic findings through complex diagnostic reasoning. We investigated whether this reason

LLT: Local Linear Transformer for PDE Operator Learning

arXiv:2607.07718v1 Announce Type: cross Abstract: Neural operators have become a common approach for learning PDE solution maps and accelerating numerical simulations. Transformer-based neural operato

LongE2V: Long-Horizon Event-based Video Reconstruction, Prediction, and Frame Interpolation with Video Diffusion Models

arXiv:2607.08770v1 Announce Type: new Abstract: Recovering high-quality video from sparse event streams is a challenging task. Regression methods often blur textures, while existing generative models

LTM: Large-scale Terrain Model for Wildfire-prone Landscapes

arXiv:2607.08711v1 Announce Type: new Abstract: Accurate 3D terrain maps are essential for emergency response when assessing wildfire hazards. However, wildfire-prone regions often span vast areas whe

MatBind: A Shared Embedding Space for Multimodal Materials Characterization

arXiv:2607.08470v1 Announce Type: new Abstract: Fully characterizing a crystalline material requires integrating heterogeneous data sources -- atomic structures, diffraction patterns, electronic densi

Maybe @GaryMarcus is on to something…

Maybe @GaryMarcus is on to something… The Branching Dendrite Secret Behind Human Cognitive Superiority | Neuroscience News Summary: Researchers have discovered that the fundamental building blocks of

safetygary-marcus--x

Mechanistic Interpretability of LLM Jailbreaks via Internal Attribution Graphs

arXiv:2607.07903v1 Announce Type: cross Abstract: Large language models (LLMs) exhibit remarkable capabilities but remain highly vulnerable to adversarial prompts and jailbreak attacks. Existing appro

MentalHospital: A Virtual Environment for Evaluating Psychiatric Clinical Encounters

arXiv:2607.08257v1 Announce Type: new Abstract: Large language models (LLMs) have shown strong performance on isolated psychiatric tasks, including dialogue, diagnosis, and treatment planning, yet exi

MPFlow: Learning Budgeted Max-Flow Optimization on the Lightning Network with Deep Graph Reinforcement Learning

arXiv:2607.08703v1 Announce Type: new Abstract: We address liquidity placement in the Bitcoin Lightning Network (LN): given a fixed budget, which channels should a node open to maximize its routing ca

Multi-Distribution Robust Conformal Prediction

arXiv:2601.02998v2 Announce Type: replace Abstract: In many fairness and distribution robustness problems, one has access to labeled data from multiple source distributions yet the test data may come

MultiFair: Multimodal Balanced Fairness-Aware Medical Classification with Dual-Level Gradient Modulation

arXiv:2510.07328v2 Announce Type: replace-cross Abstract: Medical decision systems increasingly rely on data from multiple sources to ensure reliable and unbiased diagnosis. However, existing multimod

Multimodal Unlearning Across Vision, Language, Video, and Audio: Survey of Methods, Datasets, and Benchmarks

arXiv:2607.07907v1 Announce Type: cross Abstract: With the growing adoption of VLMs, DMs, LLMs, and AFMs, these multimodal foundation models can inadvertently encode sensitive, copyrighted, biased, or

Native Video-Action Pretraining for Generalizable Robot Control

arXiv:2607.08639v1 Announce Type: cross Abstract: The advent of video-action models offers a promising path for robot control. Nevertheless, we argue that repurposing video generative models designed

NFTR: From Provable Mode-Averaging to Geodesic Subgoal Selection in Offline Goal-Conditioned RL

arXiv:2607.07855v1 Announce Type: new Abstract: Hierarchical Implicit Q-Learning (HIQL), an offline goal-conditioned RL method, selects subgoals by value-function advantages alone. This rule has two c

One of the biggest threats to US innovation isn’t just research funding - it’s losing the next generation of scientists. Phd admissions at l…

One of the biggest threats to US innovation isn’t just research funding - it’s losing the next generation of scientists. Phd admissions at leading research universities fell 15% this year. Scientific

safetygary-marcus--x

Open-ended Multi-agent Autocurricula via Visual Inspection of Policies with Multi-modal LLMs

arXiv:2607.08193v1 Announce Type: cross Abstract: Open-ended curricula in Reinforcement Learning (RL) aim to train generally-capable agents by identifying tasks that facilitate learning increasingly c

OpenAI's head of safety, Johannes Heidecke, is leaving as OpenAI integrates its research and safety teams; Mia Glaese will become VP of research and safety (Maxwell Zeff/Wired)

Maxwell Zeff / Wired: OpenAI's head of safety, Johannes Heidecke, is leaving as OpenAI integrates its research and safety teams; Mia Glaese will become VP of research and safety — Johannes Heidecke's

safetytechmeme

OPSD-V: On-Policy Self-Distillation for Post-Training Few-Step Autoregressive Video Generators

arXiv:2607.08766v1 Announce Type: new Abstract: We propose OPSD-V, an on-policy self-distillation paradigm for post-training few-step autoregressive (AR) video diffusion models. Existing few-step AR v

Persona Cartography: Charting Language Model Personality Traits in Weight Space

arXiv:2607.07916v1 Announce Type: new Abstract: Large language models exhibit recurring behavioural patterns -- personas -- that shape generalisation and safety, but we lack reliable tools for decompo

PhyMAGIC: Physical Motion-Aware Generative Inference with Confidence-guided LLM

arXiv:2505.16456v3 Announce Type: replace Abstract: Recent advances in 3D content generation have amplified demand for dynamic models that are both visually realistic and physically consistent. Howeve

Physics-Guided Biomechanical Gait Adaptation for Humanoid Locomotion on Extreme Sloped Terrains

arXiv:2607.07830v1 Announce Type: new Abstract: Model-free reinforcement learning has enabled impressive humanoid locomotion; however, control on steep slopes remains largely unexplored. Unlike flat o

PLURAL: A Global Dataset for Value Alignment

arXiv:2607.08034v1 Announce Type: cross Abstract: Large language models (LLMs) are used worldwide, yet disproportionately reflect Western values, limiting their ability to represent diverse value syst

Post-Training in End-to-End Autonomous Driving

arXiv:2607.08072v1 Announce Type: new Abstract: End-to-end models that map multimodal inputs directly to future trajectories/maneuvers have emerged as an increasingly prominent research paradigm in au

Prismata: Confining Cross-Site Prompt Injection in Web Agents

arXiv:2607.08147v1 Announce Type: cross Abstract: Autonomous web agents promise to automate everyday browsing tasks, but inherit one of the web's oldest attack surfaces. Cross-Site Scripting proved th

reminds me of the time Greg Brockman personally downloaded YouTube videos for training purposes, as reported in NYT.

reminds me of the time Greg Brockman personally downloaded YouTube videos for training purposes, as reported in NYT. 🦔Apple is suing OpenAI for systematic trade secret theft weeks before the IPO. The

safetygary-marcus--x

Safe Flow Q-Learning: Offline Safe Reinforcement Learning with Reachability-Based Flow Policies

arXiv:2603.15136v2 Announce Type: replace-cross Abstract: Offline safe reinforcement learning (RL) seeks reward-maximizing policies from static datasets under strict safety constraints. Existing metho

SASGeo: Stability-Aware Semantic Map Localization for GNSS-Denied UAVs -- A Framework and Synthetic Proof of Concept

arXiv:2607.07737v1 Announce Type: cross Abstract: GNSS-denied unmanned aerial vehicles require occasional absolute position fixes to bound the drift of visual-inertial odometry. Cross-view image retri

Scalable and Culturally Specific Stereotype Dataset Construction via Human-LLM Collaboration

arXiv:2607.07895v1 Announce Type: new Abstract: Research on stereotypes in large language models (LLMs) has largely focused on English-speaking contexts, due to the lack of datasets in other languages

safetyarxiv-cs-cl

Search-based Testing of Vision Language Models for In-Car Scene Understanding

arXiv:2607.02300v2 Announce Type: replace Abstract: In the automotive domain, in-car scene understanding (ISU) enables the detection of safety-critical events, such as driver distraction, and supports

Securing Autonomous Vehicle Systems via Twin-Aware Federated Reinforcement Learning

arXiv:2607.08137v1 Announce Type: cross Abstract: Federated reinforcement learning (FRL) is crucial for enabling collaborative learning across multiple agents without sharing raw data, thereby enhanci

SeFA-Policy: Fast and Accurate Visuomotor Policy Learning with Selective Flow Alignment

arXiv:2511.08583v2 Announce Type: replace-cross Abstract: Developing efficient and accurate visuomotor policies poses a central challenge in robotic imitation learning. While recent rectified flow app

SkillPlug: Unsupervised Skill Mining for Few-Shot Adaptation in Robotic Manipulation

arXiv:2607.08354v1 Announce Type: new Abstract: Learning transferable visuomotor imitation policies that generalize across diverse manipulation tasks and adapt rapidly to new tasks from only a handful

Soft Robotic Exogloves for Dexterous Mobility -- Towards Personalized Rehabilitation

arXiv:2607.07968v1 Announce Type: new Abstract: Soft robotic exogloves can provide hand rehabilitation and assistance. Fitting these gloves often relies on standardized measurements not tailored to th

Spatio-Temporal Scheduling Prediction Under Backhaul Delay for Resilient Coordinated Beamforming

arXiv:2607.08454v1 Announce Type: cross Abstract: Coordinated beamforming in distributed 5G networks relies on the timely exchange of inter-cell scheduling information, but backhaul latency makes this

SPHINX: First Explain, Then Explore

arXiv:2606.17482v2 Announce Type: replace Abstract: Generating adversarial driving scenarios is critical for evaluating and improving autonomous vehicle decision-making systems in simulation. Recent a

Statistical Efficiency and Inference of Quantile Distributional Reinforcement Learning

arXiv:2607.08444v1 Announce Type: cross Abstract: In this paper, we study quantile-based distributional reinforcement learning from the perspective of statistical efficiency. We focus on distributiona

Texture Representations in Deep Vision Models: Comparing CNNs, Vision Transformers, and Human Perception

arXiv:2607.08321v1 Announce Type: new Abstract: In computational vision science, Convolutional Neural Networks (CNNs) have emerged as a popular model of biological vision because of the alignment they

The Contribution of XAI for the Safe Development and Certification of AI: An Expert-Based Analysis

arXiv:2408.02379v2 Announce Type: replace-cross Abstract: Developing and certifying safe - or so-called trustworthy - AI has become an increasingly salient issue, especially in light of upcoming regul

TNODEV: Toolbox for Neural ODE Verification

arXiv:2606.16567v2 Announce Type: replace Abstract: Neural ordinary differential equations (neural ODE) gained attention in safety critical settings such as continuous-time controllers for cyber-physi

Trustworthy Machine Learning through the Lens of Combinatorial Optimization: Survey and Research Perspectives

arXiv:2607.07762v1 Announce Type: new Abstract: Modern machine learning (ML) increasingly relies on complex models whose behavior is difficult to characterize beyond empirical performance metrics. Acr

Two Axes of LLM Abstention: Answer Correctness and Question Answerability

arXiv:2607.08456v1 Announce Type: cross Abstract: A model should refuse two different things: answers it would get wrong, and questions it should not answer at all, such as unanswerable ones or ones r

UltraX: Refining Pre-Training Data at Scale with Adaptive Programmatic Editing

arXiv:2607.08646v1 Announce Type: cross Abstract: As available training data approaches its physical limit, gains from Scaling Laws have begun to diminish. Consequently, improving Large Language Model

Unified Face Attack Detection via Fine-Grained Semantic Guidance

arXiv:2607.08156v1 Announce Type: new Abstract: The growing applications of facial recognition systems are accompanied by increasingly diverse security threats. Existing datasets lack detailed textual

Validating LLMs in social science: Epistemic threats and emerging norms

arXiv:2607.07915v1 Announce Type: cross Abstract: Large language models (LLMs) are reshaping social science methodology. Researchers increasingly prompt language models to generate quantitative measur

safetyarxiv-cs-cl