safety

We also had three third-party AI safety organizations provide feedback on our analysis: @redwood_ai, @apolloaievals, @METR_Evals. You can fi…

We also had three third-party AI safety organizations provide feedback on our analysis: @redwood_ai, @apolloaievals, @METR_Evals. You can find @redwood_ai's report here: https://blog.redwoodresearch.o

safetyopenai--x

8 May 2026

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

We’ve spent a lot of time on the framework underneath Codex, so it can move quickly on routine work while stopping for review when the risk …

DGX agent

We’ve spent a lot of time on the framework underneath Codex, so it can move quickly on routine work while stopping for review when the risk changes. Here’s how we use sandboxing, approvals, network po

safetysam-altman--x

8 May 2026

3.5 years into the “AI revolution” Basically no sign that it has mattered for productivity.

Gary Marcus argues that despite 3.5 years of rapid AI development and deployment, there is minimal empirical evidence that AI has significantly impacted overall economic productivity metrics. The post

A Closed-Form Dual-Barrier CBF Safety Filter for Holonomic Robots on Incrementally Built Occupancy Grid Maps

arXiv:2605.05182v1 Announce Type: new Abstract: We present a dual-barrier control barrier function (CBF) safety filter for real-time, safety-critical velocity control of holonomic robots operating in

safetyarxiv-cs-ro

A cross-modal network for facial expression recognition

arXiv:2605.04439v1 Announce Type: new Abstract: Deep neural networks enriched with structural information have been widely employed for facial expression recognition tasks. However, these methods ofte

A Skill-Based AI Agentic Pipeline for Library of Congress Subject Indexing

arXiv:2605.03537v1 Announce Type: cross Abstract: This paper presents a modular AI agentic skill pipeline for automating subject indexing with Library of Congress Subject Headings (LCSH). Subject inde

Adapt to Thrive! Adaptive Power-Mean Policy Optimization for Improved LLM Reasoning

arXiv:2605.04066v1 Announce Type: new Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) is an essential paradigm that enhances the reasoning capabilities of Large Language Models (LLMs).

Adaptive Dual-Path Framework for Covert Semantic Communication

arXiv:2605.03423v1 Announce Type: new Abstract: This paper proposes a novel adaptive dual-path framework for covert semantic communication (SemCom), which integrates covert information transmission wi

Adaptive Policy Selection and Fine-Tuning under Interaction Budgets for Offline-to-Online Reinforcement Learning

arXiv:2605.05123v1 Announce Type: new Abstract: In offline-to-online reinforcement learning (O2O-RL), policies are first safely trained offline using previously collected datasets and then further fin

Advancing Analytic Class-Incremental Learning through Vision-Language Calibration

arXiv:2602.13670v2 Announce Type: replace Abstract: Class-incremental learning (CIL) with pre-trained models (PTMs) faces a critical trade-off between efficient adaptation and long-term stability. Whi

Agent-Based Modeling of Low-Emission Fertilizer Adoption for Dairy Farm Decarbonisation using Empirical Farm Data

arXiv:2605.03648v1 Announce Type: new Abstract: To understand complex system dynamics in dairy farming, it is essential to use modeling tools that capture farm heterogeneity, social interactions, and

Anatomy of a failure: When, how, and why deep vision fails in scientific domains

arXiv:2605.04231v1 Announce Type: new Abstract: Mirroring its ubiquity in popular media and all human activities, the use of deep learning (DL) is rapidly growing in scientific imaging modalities. How

Anticipating Innovation Using Large Language Models

arXiv:2605.04875v1 Announce Type: new Abstract: Forecasting innovation, intended as the emergence of new technological combinations, is a fundamental challenge for science and policy. We show that for

AnyPos: Automated Task-Agnostic Actions for Bimanual Manipulation

arXiv:2507.12768v2 Announce Type: replace Abstract: Learning generalizable manipulation policies hinges on data, yet robot manipulation data is scarce and often entangled with specific embodiments, ma

*apologies for typo: 300 not 30 megawatts

Gary Marcus corrected a previous statement about power consumption, clarifying that the figure was 300 megawatts rather than 30 megawatts. This appears to be a technical correction posted on X regardi

Balanced Aggregation: Understanding and Fixing Aggregation Bias in GRPO

arXiv:2605.04077v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) has become a central paradigm for improving reasoning and code generation in large language mode

Beyond Fixed Thresholds and Domain-Specific Benchmarks for Explainable Multi-Task Classification in Autonomous Vehicles

arXiv:2605.04299v1 Announce Type: new Abstract: Scene understanding is a vital part of autonomous driving systems, which requires the use of deep learning models. Deep learning methods are intrinsical

Beyond Public Access in LLM Pre-Training Data

arXiv:2505.00020v2 Announce Type: replace Abstract: Using a legally obtained dataset of 34 copyrighted O'Reilly Media books, we apply the DE-COP membership inference attack method to investigate wheth

Brainrot: Deskilling and Addiction are Overlooked AI Risks

arXiv:2605.03512v1 Announce Type: cross Abstract: The scope of AI safety and alignment work in generative artificial intelligence (GenAI) has so far mostly been limited to harms related to: (a) discri

Causal discovery under mean independence and linearity

arXiv:2605.04381v1 Announce Type: cross Abstract: Causal discovery methods such as LiNGAM identify causal structure from observational data by assuming mutually independent disturbances. This assumpti

ChatGPT’s ‘Trusted Contact’ will alert loved ones of safety concerns

OpenAI is launching an optional safety feature for ChatGPT that allows adult users to assign an emergency contact for mental health and safety concerns. Friends, family members, or caregivers designat

safetythe-verge-ai

CHE-TKG: Collaborative Historical Evidence and Evolutionary Dynamics Learning for Temporal Knowledge Graph Reasoning

arXiv:2605.04652v1 Announce Type: new Abstract: Temporal knowledge graph (TKG) reasoning aims to predict future events from historical facts. A key challenge lies in jointly capturing two sources of p

Conditional Flow-VAE for Safety-Critical Traffic Scenario Generation

arXiv:2605.04366v1 Announce Type: cross Abstract: Safety-critical scenarios are essential for the development of autonomous vehicles (AVs) but are rare in real-world driving data. While simulation off

Confronting Label Indeterminacy in Automated Bail Decisions

arXiv:2605.04073v1 Announce Type: new Abstract: Bail decisions present a fundamental challenge for data-driven decision support systems. When bail is denied, the counterfactual outcome of whether the

Connecting online criminal behavior with machine learning: Using authorship attribution to analyze and link potential online traffickers

arXiv:2605.04080v1 Announce Type: new Abstract: This research investigated how online criminal activities can be better understood and connected using data-driven machine learning methods. Many illega

contest time:

contest time: First person to update this chart with no mistakes using an AI system and a general prompt (rather than one than that handholds the system every step of the win) wins a signed book! Plea

Copula-Based Endogeneity Correction for Doubly Robust Estimation of Treatment Effect

arXiv:2605.03278v2 Announce Type: cross Abstract: Doubly Robust (DR) estimation of treatment effect relies on an untestable assumption that is the absence of unobserved confounding. This assumption is

Counter-Dyna: Data-Efficient RL-Based HVAC Control using Counterfactual Building Models

arXiv:2605.04555v1 Announce Type: new Abstract: Model-based reinforcement learning (MBRL) offers a promising approach for data-efficient energy management in buildings, combining the strengths of pred

CRAFT: Counterfactual-to-Interactive Reinforcement Fine-Tuning for Driving Policies

arXiv:2605.04470v1 Announce Type: new Abstract: Open-loop imitation learning has advanced modern autonomous driving policy architectures, but closed-loop deployment remains vulnerable to policy-induce

D-OPSD: On-Policy Self-Distillation for Continuously Tuning Step-Distilled Diffusion Models

arXiv:2605.05204v1 Announce Type: new Abstract: The landscape of high-performance image generation models is currently shifting from the inefficient multi-step ones to the efficient few-step counterpa

Data-dependent Exploration for Online Reinforcement Learning from Human Feedback

arXiv:2605.04477v1 Announce Type: new Abstract: Online reinforcement learning from human feedback (RLHF) has emerged as a promising paradigm for aligning large language models (LLMs) by continuously c

Decompose to Understand, Fuse to Detect: Frequency-Decoupled Anomaly Detection for Encrypted Network Traffic

arXiv:2605.02970v1 Announce Type: cross Abstract: Network traffic anomaly detection represents a critical cybersecurity task, yet widespread encryption makes this task increasingly challenging. In res

DFPO: Scaling Value Modeling via Distributional Flow towards Robust and Generalizable LLM Post-Training

arXiv:2602.05890v2 Announce Type: replace-cross Abstract: Training reinforcement learning (RL) systems in real-world environments remains challenging due to noisy supervision and poor out-of-domain (O

Direct Product Flow Matching: Decoupling Radial and Angular Dynamics for Few-Shot Adaptation

arXiv:2605.05054v1 Announce Type: new Abstract: Recent flow matching (FM) methods improve the few-shot adaptation of vision-language models, by modeling cross-modal alignment as a continuous multi-ste

Discovering Sparse Counterfactual Factors via Latent Adjustment for Survey-based Community Intervention

arXiv:2605.04460v1 Announce Type: new Abstract: Transportation surveys are widely used to understand travel preferences and adoption barriers, yet most survey-based analyses remain descriptive or pred

Distilling Bayesian Belief States into Language Models for Auditable Negotiation

arXiv:2605.04507v1 Announce Type: new Abstract: Negotiation agents must infer what their counterpart values, update those beliefs over dialogue turns, and choose actions under uncertainty. End-to-end

Dream-MPC: Gradient-Based Model Predictive Control with Latent Imagination

arXiv:2605.04568v1 Announce Type: new Abstract: State-of-the-art model-based Reinforcement Learning (RL) approaches either use gradient-free, population-based methods for planning, learned policy netw

Dynamic Hyperparameter Importance for Efficient Multi-Objective Optimization

arXiv:2601.03166v2 Announce Type: replace Abstract: Choosing a suitable ML model is a complex task that can depend on several objectives, e.g., accuracy, fairness, or energy consumption. In practice,

Efficiency of Parallel and Restart Exploration Strategies in Model Free Stochastic Simulations

arXiv:2503.03565v3 Announce Type: replace-cross Abstract: We analyze the efficiency of parallelization and restart mechanisms for stochastic simulations in model-free settings, where the underlying sy

Efficient Geometry-Controlled High-Resolution Satellite Image Synthesis

arXiv:2605.04557v1 Announce Type: new Abstract: High-resolution satellite images are often scarce and costly, especially for remote areas or infrequent events. This shortage hampers the development an

Efficient Model-Based Reinforcement Learning for Robot Control via Online Optimization

arXiv:2510.18518v2 Announce Type: replace Abstract: We present an online model-based reinforcement learning algorithm suitable for controlling complex robotic systems directly in the real world. Unlik

safetyarxiv-cs-ro

Efficiently Aligning Language Models with Online Natural Language Feedback

arXiv:2605.04356v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards has been used to elicit impressive performance from language models in many domains. But, broadly benefic

Elicitation Matters: How Prompts and Query Protocols Shape LLM Surrogates under Sparse Observations

arXiv:2605.04764v1 Announce Type: new Abstract: Large language models are increasingly used as surrogate models for low-data optimization, but their optimizer-facing prediction and its uncertainty rem

Elon Musk’s DOGE “blatantly used” race, gender and other protected characteristics to execute the largest mass termination of federal grants…

Elon Musk’s DOGE “blatantly used” race, gender and other protected characteristics to execute the largest mass termination of federal grants in the history of the National Endowment for the Humanities

Encoding Predictability and Legibility for Style-Conditioned Diffusion Policy

arXiv:2603.16368v2 Announce Type: replace-cross Abstract: Striking a balance between efficiency and transparent motion is a core challenge in human-robot collaboration, as highly expressive movements

Enhancing the interpretability of spatially variable N2O model predictions with soft sensors during wastewater treatment

arXiv:2605.04082v1 Announce Type: new Abstract: Model-based solutions for nitrous oxide (N2O) emissions from wastewater treatment plants (WWTP) are informed by operational datasets designed to control

EO-1 is now available directly in LeRobot! You can train, evaluate, and deploy EO-1 through the standard LeRobot policy interface, making it…

EO-1 is now available directly in LeRobot! You can train, evaluate, and deploy EO-1 through the standard LeRobot policy interface, making it much easier to try EO-1 on robot-control workflows. Huge th

safetyclem-delangue--x

EP-GRPO: Entropy-Progress Aligned Group Relative Policy Optimization with Implicit Process Guidance

arXiv:2605.04960v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR), particularly Group Relative Policy Optimization (GRPO), has advanced LLM reasoning. However, GRPO