Search: “safety”

Watch the goal post shift unfold in real time: AGI used to be doing anything a person, including an expert, could do, and by the end of the …

Watch the goal post shift unfold in real time: AGI used to be doing anything a person, including an expert, could do, and by the end of the interview it’s not. I call this the AI bait and switch. As I

Safety

We are saying no to a data centre in Île-des-Chênes because there are big threats to the environment and not much benefit to the economy. Ou…

DGX agent

We are saying no to a data centre in Île-des-Chênes because there are big threats to the environment and not much benefit to the economy. Our message to any tech company out there… if you want to buil

safetygary-marcus--x

4 Jun 2026

What Type of Inference is Active Inference?

arXiv:2606.04935v1 Announce Type: new Abstract: Active inference casts decision-making as inference, with the Expected Free Energy (EFE) unifying goal-directed and information-seeking behavior. Recent

When Both Layers Learn: Training Dynamics of Representing Linear Models via ReLU Networks

arXiv:2606.04476v1 Announce Type: new Abstract: In this paper, we study the gradient descent dynamics for jointly training both layers of a one-hidden-layer ReLU network to fit a linear target functio

X4Val: Learning Neural Surrogates for Variance-Reduced Policy Evaluation

arXiv:2606.05159v1 Announce Type: new Abstract: Rigorous evaluation of learning-based robotic systems is an essential prerequisite for deployment. However, real-world test data is expensive to gather;

You know this is about Sam from line 1, even before you get to the picture.

You know this is about Sam from line 1, even before you get to the picture. TL;DR: He’s going to keep bullshitting his way to IPO. This entire industry is based on an illusion. It deliberately mistake

ZeroWBC: Learning Natural Whole-Body Humanoid Interaction from Human Egocentric Data

arXiv:2603.09170v2 Announce Type: replace-cross Abstract: Achieving versatile and natural whole-body humanoid interaction control remains challenging due to the high cost of whole-body teleoperation d

A blueprint for democratic governance of frontier AI

This OpenAI publication outlines a proposed framework for establishing democratic oversight and governance structures for advanced AI systems, addressing how frontier AI development should be regulate

safetyopenai

A Cartesian-3j Framework for Machine Learning Interatomic Potentials

arXiv:2512.16882v2 Announce Type: replace-cross Abstract: Machine learning interatomic potentials (MLIPs) have brought substantial gains in the extrapolation capability in computational chemistry. How

A Negative Result on Cross-Model Activation Transfer in a Pythia Multi-Hop Setting

arXiv:2606.03280v1 Announce Type: new Abstract: Recent work shows that language models can transmit behavioural traits through hidden signals in generated data during training. We ask whether a more d

A Pocket Offline Model for Simultaneous Speech Translation as CUNI Submission to IWSLT 2026

arXiv:2606.03948v1 Announce Type: new Abstract: We implement simultaneous translation capability with the offline direct speech-to-text translation model Canary, using the state-of-the-art policy Alig

“a train wreck’ - @edels0n Not worth half of what you will be paying – Morningstar At most, worth half of what you will be paying – Martin P…

“a train wreck’ - @edels0n Not worth half of what you will be paying – Morningstar At most, worth half of what you will be paying – Martin Peers, The Information “financial terrorism” - @egrefen Get y

Adaptive Causal Alignment for High-Confidence Adversarial Training

arXiv:2606.03925v1 Announce Type: new Abstract: Inverse adversarial training leverages high-confidence predictions to stabilize robust learning, yet we uncover a critical paradox: high confidence ofte

safetyarxiv-cs-cv

AirDreamer: Generalist Drone Navigation with World Models

arXiv:2606.03252v1 Announce Type: cross Abstract: Navigating a drone in unseen and cluttered environments requires reliable generalization to unseen scene layouts and understanding of environmental st

Aletheia: What Makes RLVR For Code Verifiers Tick?

arXiv:2601.12186v3 Announce Type: replace-cross Abstract: Multi-domain thinking verifiers trained via Reinforcement Learning with Verifiable Rewards (RLVR) are a cornerstone of modern post-training. H

Align-KD: Distilling Cross-Modal Alignment Knowledge for Mobile Vision-Language Model Enhancement

arXiv:2412.01282v2 Announce Type: replace-cross Abstract: Vision-Language Models (VLMs) bring powerful understanding and reasoning capabilities to multimodal tasks. Meanwhile, the great need for capab

Aligning Data-Driven Predictors with Allocation: A Decision-Focused Approach to Survival Analysis

arXiv:2606.02671v1 Announce Type: cross Abstract: Machine learning predictors have become essential tools for guiding automated decision making. However, a major misalignment persists: predictive mode

Alignment-Aware Decoding

arXiv:2509.26169v2 Announce Type: replace Abstract: Alignment of large language models remains a central challenge in natural language processing. Preference optimization has emerged as a popular and

an astonishing statistic, consistent with the claim I made the other day that Elon’s best days may be behind him:

Gary Marcus shared a statistic on X that he claims supports his earlier assertion that Elon Musk's most successful period may be in the past. The post likely presents data related to Musk's business p

Are we really tilting? The mechanics of reward guidance in flow and diffusion models

arXiv:2606.02884v1 Announce Type: cross Abstract: Reward guidance algorithms steer a learned generative process toward the reward-tilted measure at inference time. While empirically powerful, these me

ASAP: Exploiting the Satisficing Generalization Edge in Neural Combinatorial Optimization

arXiv:2501.17377v4 Announce Type: replace-cross Abstract: Deep Reinforcement Learning (DRL) has emerged as a promising approach for solving Combinatorial Optimization (CO) problems, such as the 3D Bin

ASymPO: Asymmetric-Scale Policy Optimization for Asynchronous LLM Post-Training Without Behavior Information

arXiv:2606.03070v1 Announce Type: cross Abstract: Asynchronous reinforcement learning can improve language-model post-training throughput by decoupling response generation from policy optimization, bu

Attention Calibration for Position-Fair Dense Information Retrieval

arXiv:2606.02737v1 Announce Type: cross Abstract: Dense retrieval models exhibit positional bias: retrieval effectiveness degrades when relevant information appears later in a passage (Zeng et al., 20

Bayesian Tensor Decomposition with Diffusion Model Prior

arXiv:2606.03212v1 Announce Type: new Abstract: Low-rank tensor decomposition (TD) is usually effective on clean, fully observed data, but it often degrades under severe missingness or noise. Low-rank

Best of Both Worlds: Multimodal Reasoning and Generation via Unified Discrete Flow Matching

arXiv:2602.12221v2 Announce Type: replace Abstract: We propose UniDFlow, a unified discrete flow-matching framework for multimodal understanding, generation, and editing. It decouples understanding an

safetyarxiv-cs-cv

Bionic Human-Motion Style Transfer for Physically Executable Whole-Body Control of Humanoid Robots

arXiv:2606.03536v1 Announce Type: new Abstract: Expressive whole-body motion is important for humanoid robots operating in human environments, where robots are expected to move stably while presenting

Black-box, Adaptive, Efficient, Transferable, Harmful, Applicable... Attacks Are All You Need to Break LLMs

arXiv:2606.03647v1 Announce Type: cross Abstract: Accurately evaluating adversarial robustness is a longstanding challenge. A flawed attack design can inflate robustness estimates, making deployment r

Breaking the Self-Confirming Loop: Diagnosing and Mitigating Systemic Reward Bias in Self-Rewarding RL

arXiv:2510.08977v2 Announce Type: replace-cross Abstract: Reinforcement learning with verifiable rewards (RLVR) efficiently scales the reasoning ability of large language models (LLMs) but is bottlene

Brief Announcement: Generative Markov Model for Distributed Computing Systems

arXiv:2606.03061v1 Announce Type: cross Abstract: Emerging distributed computing paradigms, such as the computing continuum, are inherently heterogeneous, stochastic, and complex. Efficiently and effe

Building Better Activation Oracles

arXiv:2606.02609v1 Announce Type: cross Abstract: Activation Oracles (AOs) are promising methods for interpreting residual stream activations. However, current AOs face important issues, such as hallu

Coherence Maximization Improves Pluralistic Alignment

arXiv:2606.03110v1 Announce Type: new Abstract: Aligning AI systems with diverse human values requires value specifications grounded in concrete examples, but generating such examples without extensiv

Collab-REC: An LLM-based Agentic Framework for Balancing Recommendations in Tourism

arXiv:2508.15030v5 Announce Type: replace Abstract: We propose COLLAB-REC, a multi-agent framework designed to counteract popularity bias and improve diversity in tourism recommendations. In our setup

Consistency Training Can Entrench Misalignment

arXiv:2606.03810v1 Announce Type: cross Abstract: Consistency training encourages a model to produce similar outputs across related inputs or sampling procedures. Such methods are simple, scalable, an

ConTrack: Constrained Hand Motion Tracking with Adaptive Trade-off Control

arXiv:2606.03177v1 Announce Type: new Abstract: Human demonstrations provide strong priors for robot manipulation, yet it is non-trivial to transfer them to execute on real robots due to the kinematic

ConTraIRL: Factorized Contrastive Abstractions for Transferable IRL

arXiv:2606.03017v1 Announce Type: cross Abstract: Reward transfer in Inverse Reinforcement Learning (IRL) is unreliable when policies must generalize to unseen combinations of environment dynamics and

Contrastive Neural Algorithmic Reasoning for Graph Coloring

arXiv:2606.03923v1 Announce Type: new Abstract: Graph coloring seeks to assigns colors to a graph's nodes so that adjacent nodes receive different colors, using as few colors as possible. Here, we stu

Correcting Neural Operator Spectral Bias via Diffusion Posterior Sampling with Sparse Observations

arXiv:2606.03936v1 Announce Type: new Abstract: Neural operator surrogates (NO) approximate PDE solutions orders of magnitude faster than numerical solvers, but suffer from spectral bias: high-frequen

CP-Agent: Context-Aware Multimodal Reasoning for Cellular Morphological Profiling under Chemical Perturbations

arXiv:2606.03435v1 Announce Type: new Abstract: Cell Painting combines multiplexed fluorescent staining, high-content imaging, and quantitative analysis to generate high-dimensional phenotypic readout

Curriculum-Adapted Robust Reinforcement Learning for UAV Deconfliction in Adversarial Environments

arXiv:2506.21129v2 Announce Type: replace-cross Abstract: Autonomous unmanned aerial vehicles (UAVs) increasingly rely on reinforcement learning (RL) for navigation. However, global navigation satelli

Data- and Variance-dependent Regret Bounds for Online Tabular MDPs

arXiv:2602.01903v2 Announce Type: replace Abstract: This work studies online episodic tabular Markov decision processes (MDPs) with known transitions and develops best-of-both-worlds algorithms that a

Denoise First, Orthogonalize Later: Understanding Momentum in Muon via Spectral Filtering

arXiv:2606.03899v1 Announce Type: new Abstract: Muon has recently demonstrated strong empirical performance in large language model training, but the theoretical role of momentum in Muon remains uncle

Denoising Tells When to Replan: Denoising-Variance Adaptive Chunking for Flow-Based Robot Policies

arXiv:2606.03847v1 Announce Type: new Abstract: Action chunking has become a common inference strategy for flow-based robot policies, improving action coherence by modeling multi-step temporal depende

Discovering autonomous quantum error correction via deep reinforcement learning

arXiv:2511.12482v2 Announce Type: replace-cross Abstract: Quantum error correction is essential for fault-tolerant quantum computing. However, standard methods relying on active measurements may intro

Do Explanations Increase the Risk of Decision Logic Leakage? Explanation-Guided Stealing of Graph Models

arXiv:2506.03087v2 Announce Type: replace-cross Abstract: Graph Neural Networks (GNNs) have become essential tools for analyzing graph-structured data in domains such as drug discovery and financial a

Do Neural Retrievers Prefer Certain Documents? Evidence of Learned Relevance Priors

arXiv:2606.02814v1 Announce Type: cross Abstract: Neural retrievers are trained to estimate query-document relevance from annotated query-document pairs. Yet annotation protocols may not purely reflec

DriftSched: Adaptive QoS-Aware Scheduling under Runtime Token Drift for Multi-Tenant GPU Inference

arXiv:2606.02982v1 Announce Type: cross Abstract: The rapid growth of large language model (LLM) inference services has increased the demand for efficient multi-tenant GPU scheduling. While modern inf

Dynamic Short Convolutions Improve Transformers

arXiv:2606.03825v1 Announce Type: cross Abstract: Transformers have become the dominant architecture for large language models, largely due to the scalability and flexibility of attention, feed-forwar

Effect of Demographic Bias on Skin Lesion Classification

arXiv:2606.03214v1 Announce Type: new Abstract: In this study, we evaluate the performance of skin lesion classification using ResNet-based convolutional models, focusing on the impact of demographic