safety

arXiv:2607.11953v1 Announce Type: new Abstract: Does a reinforcement-learning agent that earns high reward represent its task's latent state, or only a reward-correlated shortcut? The question is usua

arXiv:2607.12790v1 Announce Type: new Abstract: Self-evolving agent systems improve by creating, revising, and retiring their own skills, but every such loop rests on a hidden assumption: a reliable e

little by little, OpenAI’s storytelling is falling apart. my 2023 projection that they would someday be viewed as the WeWork of AI is looking stronger by the day. OpenAI is on pace to miss its own fiv

New work led by @FlemmingKondrup and @tomjiralerspong highlights an important vulnerability in chain-of-thought monitoring for agents, I highly recommend giving it a read. Link to the paper: https://a

In this post, you will learn how ScienceSoft, an Amazon Web Services (AWS) Services Partner, integrated Amazon Nova 2 Sonic with Amazon Bedrock Guardrails to build a Health Insurance Portability and A

Absolutely fascinating work by @SakanaAILabs reproducing @kenneth0stanley Picbreeder in a non-interactive, VLM-agentic way. I've had years to reflect on Kenneth Stanley's ideas as originally communica

For almost two decades people like @YLeCun and @geoffreyhinton dumped on me for saying we need symbols in addition to deep learning. But that’s exactly what loop engineering is: adding symbols to deep

The Anglo-Scottish Enlightenment – the real antidote to Rousseau and Voltaire The French Enlightenment and the Anglo-Scottish Enlightenment happened simultaneously, in the same century, reading the sa

Gary Marcus argues that the engineering and infrastructure surrounding large language models are more critical to their practical success than the models themselves. This reflects his broader perspect

arXiv:2511.03075v2 Announce Type: replace Abstract: The safe deployment of autonomous systems in safety-critical settings requires a paradigm that combines human expertise with AI-driven analysis, esp

arXiv:2607.08196v1 Announce Type: new Abstract: As part of a series on first-principles modeling of cognitive functions, this paper attempts to provide a mathematical formulation of thinking and perce

Noam Scheiber / New York Times: A US NLRB judge rules that Atlassian had illegally fired an employee in 2023 for pushing back against manager layoffs, and orders reinstatement and compensation — A fed

arXiv:2607.08443v1 Announce Type: cross Abstract: Dynamic traffic variations in Open Radio Access Networks (O-RAN) lead to drift, which degrades the performance of Artificial Intelligence/Machine Lear

Ahead of a dinner with a US senator, AI researcher Nate Soares (@So8res) was told: 'Don't give them any of the crazy crap. You know, play it cool.' His friends opened with the concern that someone cou

arXiv:2607.08043v1 Announce Type: cross Abstract: Research software collaborations span meetings, informal chats, pull requests, and GitHub issues. A decision surfaced in a Slack thread, refined in a

arXiv:2607.07766v1 Announce Type: new Abstract: Large language models (LLMs) have become significant providers of mental health support, yet they remain products of an attention economy whose operatio

As part of our ongoing efforts to strengthen our safeguards for advanced AI capabilities in biology, we’re evolving our Bio Bug Bounty into an ongoing private program, known as the OpenAI Bio Bug Boun

arXiv:2607.08335v1 Announce Type: cross Abstract: Policy-based approaches to Bayesian experimental design (BED) allow the learning of deep policy networks that adaptively make intelligent design decis

arXiv:2607.08256v1 Announce Type: cross Abstract: Best-of-N (BoN) inference improves content consistency in zero-shot text-to-speech by selecting from N candidates with an automatic speech recognition

arXiv:2602.05459v2 Announce Type: replace Abstract: Offline goal-conditioned reinforcement learning (GCRL) is typically benchmarked by the best tuned success rate of each method. This score measures a

arXiv:2512.15138v2 Announce Type: replace Abstract: Reference-guided instance editing is fundamentally limited by semantic entanglement, where a reference's intrinsic appearance is intertwined with it

breaking: company built on stolen IP and lies allegedly steals more IP “OpenAI’s nascent hardware business now rests on the shakiest of foundations, rotten to its core by its illegal reliance on misap

arXiv:2601.11073v3 Announce Type: replace-cross Abstract: Online financial services constitute an essential component of contemporary web ecosystems, yet their openness introduces substantial exposure

arXiv:2607.08555v1 Announce Type: new Abstract: The operational integrity of complex industrial systems relies on precise anomaly detection and diagnosis. The vast majority of existing methods narrowl

arXiv:2607.08742v1 Announce Type: new Abstract: Keypoint tracking alone is insufficient for object interaction tasks such as sitting on a chair, wiping a board, or pushing furniture, where the robot c

arXiv:2607.08561v1 Announce Type: new Abstract: A series of results from the NeuroAI over the past fifteen years have raised core questions both about how to compare Deep Neural Network (DNN) models t

At Google Cloud, we take our role in the financial ecosystem very seriously. We firmly believe that operational resilience is essential to driving and sustaining responsible innovation. Today, we mark

arXiv:2602.17686v4 Announce Type: replace-cross Abstract: Distilling Chain-of-Thought (CoT) reasoning from large language models into compact student models presents a fundamental challenge: teacher r

arXiv:2511.16920v2 Announce Type: replace Abstract: Anomaly generation is often framed as few-shot fine-tuning with anomalous samples, which contradicts the scarcity that motivates generation and tend

arXiv:2607.08417v1 Announce Type: new Abstract: A Ladder Logic Bomb (LLB) is malicious control logic in a Programmable Logic Controller (PLC) program that lies dormant until a trigger activates a payl

arXiv:2511.19032v2 Announce Type: replace Abstract: Visual corruptions can change vision--language model (VLM) behavior in ways that top-1 accuracy does not capture. A model may keep the same answer w

arXiv:2607.08031v1 Announce Type: cross Abstract: The dynamics of communication environments induce significant distribution shifts across domains, challenging the generalization of deep learning-base

arXiv:2601.10504v2 Announce Type: replace Abstract: As Large Language Models (LLMs) increasingly operate as Deep Research (DR) Agents capable of autonomous investigation and information synthesis, rel

arXiv:2607.08404v1 Announce Type: cross Abstract: Current computational approaches for drug design typically focus on generating molecules conditioned on specific targets or general molecular properti

arXiv:2504.07856v4 Announce Type: replace Abstract: Curriculum learning enhances Direct Preference Optimization (DPO) for aligning Large Language Models (LLMs), yet existing methods rely on a one-dime

arXiv:2607.08504v1 Announce Type: cross Abstract: Most cooperative Vision-Language Navigation (VLN) methods assume unlimited communication, not considering real-world applications where bandwidth is r

arXiv:2603.23667v2 Announce Type: replace-cross Abstract: We introduce Echoes, a new dataset for music deepfake detection designed for training and benchmarking detectors under realistic and provider-

arXiv:2306.12857v3 Announce Type: replace Abstract: The storage, management, and application of massive spatio-temporal data are widely used in practical scenarios, including public safety. However, d

arXiv:2607.07918v1 Announce Type: cross Abstract: Current safety methods for large language models are known to be vulnerable to adversarial attacks, motivating research into robust alternatives. Late

arXiv:2607.08436v1 Announce Type: cross Abstract: Egocentric human data offers scalable supervision for robot manipulation. However, behavior cloning entangles transferable content like objects, scene

arXiv:2607.08493v1 Announce Type: cross Abstract: Subjective NLP tasks often exhibit systematic annotator disagreement, requiring models that represent uncertainty rather than collapse it. We introduc

The European Union has tentatively found that Meta Platforms Inc. breached the bloc’s DSA tech industry law. The European Commission, the EU’s executive arm, published its conclusions today. The DSA,

arXiv:2607.07967v1 Announce Type: cross Abstract: Diffusion-based policies have recently emerged as powerful policy parameterizations for reinforcement learning, representing state-conditioned action

arXiv:2607.07859v1 Announce Type: new Abstract: Reinforcement learning (RL) research has increasingly shifted focus towards alignment, ensuring agents learn behaviors adhering to human values. While h

arXiv:2607.08028v1 Announce Type: new Abstract: Enterprise large language model (LLM) applications often begin as prototypes whose behavior is carried by prompts and retrieval context. Productization

arXiv:2511.18940v3 Announce Type: replace Abstract: Cross-subject motor imagery decoding remains a fundamental challenge in EEG-based brain-computer interfaces due to substantial inter-subject variabi