safety

Let’s build seven trillion dollars worth of data centers for sports betting!

Let’s build seven trillion dollars worth of data centers for sports betting! Statistics of week: “More Americans pay for sports betting apps (5%) than pay for AI… 37% of consumers say none of AI's use

Safety

MELLA: Bridging Linguistic Capability and Cultural Groundedness for Low-Resource Language MLLMs

DGX agent

arXiv:2508.05502v2 Announce Type: replace-cross Abstract: Multimodal Large Language Models (MLLMs) perform strongly in high-resource languages, yet often produce fluent but culturally 'thin' descripti

safetyarxiv-cs-ai

24 Jul 2026

Multivariate Planar Curves: A Statistical Framework for Shape Analysis in Images

arXiv:2508.11780v3 Announce Type: replace-cross Abstract: Recent developments in computer vision have made segmented images widely available across many domains, such as medicine, where segmented radi

NVIDIA, Palantir, Replit, Microsoft, Crowdstrike, Dell and others send a strong message to congress to keep open access to open weights mode…

NVIDIA, Palantir, Replit, Microsoft, Crowdstrike, Dell and others send a strong message to congress to keep open access to open weights models. 'Our AI leadership will be judged not by one frontier AI

safetyyann-lecun--x

One Round Is All You Need: Analytic Federated Learning for Task-Heterogeneous Multi-Label Medical Image Classification

arXiv:2607.20641v1 Announce Type: new Abstract: Federated learning (FL) enables multiple clinical institutions to collaboratively train a shared disease classifier without centralizing patient data. I

Open models for the win!

Open models for the win! For my first post, I’m sharing a letter @NVIDIA signed on why open models matter. AI will transform every industry, power every company, and be built by every country. Open mo

safetyclem-delangue--x

Open weight models will ensure that the entire world benefits from AI growth, and that America does not get left behind

Open weight models will ensure that the entire world benefits from AI growth, and that America does not get left behind For my first post, I’m sharing a letter @NVIDIA signed on why open models matter

safetyarthur-mensch--x

OpenAI and Anthropic are now competing to see who can make the scariest AI for marketing purposes. What could possibly go wrong?

Pedro Domingos noted that OpenAI and Anthropic are competing to develop the scariest AI for marketing, raising concerns about the potential misuse of persuasive machine‑learning models. His comment hi

Operational Identity: A Finite Audit of Declared and Implemented Rules of Sameness

arXiv:2607.20729v1 Announce Type: cross Abstract: A record system declares when two records refer to the same entity, occurrence, scope, or rule. Its disclosed implementation mechanisms induce a corre

OPOD: On-Policy Omni Distillation

arXiv:2607.20918v1 Announce Type: new Abstract: Omni-modal models can handle text, images, and audio in one system, but improving all of these abilities together remains difficult. Training a single m

PATS: Policy-Aware Training Scaffolding for Agentic Reinforcement Learning

arXiv:2607.21419v1 Announce Type: new Abstract: In long-horizon LLM agent reinforcement learning, weak policies often repeat similar failures, producing uninformative rollout trajectories and limiting

Perspective Latents as an Architectural Condition for Causal Emergence in Active Inference Agents

arXiv:2607.20708v1 Announce Type: new Abstract: A recent line of work measures causal emergence in reinforcement learning agents through Integrated Information Decomposition, reporting that Phi_r grow

Phonetic forced alignment for low-resource language varieties: Model training and evaluation on Chengdu Mandarin

arXiv:2607.21332v1 Announce Type: cross Abstract: Phonetic forced alignment is a key technique in phonetic research, yet existing alignment systems lack specialized models for low-resource language va

PILD: Physics-Informed Learning via Diffusion

arXiv:2601.21284v2 Announce Type: replace-cross Abstract: Diffusion models have emerged as powerful generative tools for modeling complex data distributions, yet their purely data-driven nature limits

Probably no LLM will ever achieve that, no matter how many data centers they build.

Probably no LLM will ever achieve that, no matter how many data centers they build. Lets compare Amazon Prime to AI: According to market research from Consumer Intelligence Research Partners (CIRP), t

QATMA: Quantization-Aware Training with Multimodal Alignment for Open-Vocabulary Object Detection

arXiv:2603.05964v3 Announce Type: replace Abstract: Quantizing open-vocabulary object detection (OVOD) models reduces their memory and computational costs, but extremely low-bit quantization severely

Reexamining zero-shot summarization: Empirical investigation of trustworthiness of LLM-summarizers

arXiv:2607.21010v1 Announce Type: new Abstract: Zero-shot summarization using Large Language Models (LLMs) has significantly advanced the abstractive summarization task by producing coherent and fluen

Regulating autonomous and agentic AI

arXiv:2607.21345v1 Announce Type: new Abstract: Regulating activities where regulatees use autonomous and agentic AI is challenging. Regulatory assumptions about regulatee knowledge and control no lon

“related incidents have been happening for a while” and OpenAI has no real solution in sight. Confirms my longstanding conjecture that curre…

“related incidents have been happening for a while” and OpenAI has no real solution in sight. Confirms my longstanding conjecture that current approaches cannot be made safe. An OpenAI staffer talked

Reliability-Aware LLM Alignment from Inconsistent Human Feedback

arXiv:2607.20515v1 Announce Type: new Abstract: Reinforcement Learning from Human Feedback (RLHF) is critical for aligning Large Language Models (LLMs) with human preferences. However, its efficacy is

Representing Entity Importance in AI Knowledge Systems: A Dual-Signal Framework of Audience Evaluation and Structural Authority

arXiv:2607.20925v1 Announce Type: new Abstract: AI knowledge systems require representations of entity importance for retrieval, recommendation, evidence selection, and knowledge-intensive reasoning.

Respectfully, @davidsacks, I strongly disagree with your take, and I feel that you reached your conclusions without looking at the data, and…

Respectfully, @davidsacks, I strongly disagree with your take, and I feel that you reached your conclusions without looking at the data, and that your conclusions will give Americans false comfort. -

RL-MACRO: A Cybernetic Closed-Loop Intelligence Framework for Multimodal Adaptive Robotic Craniotomy

arXiv:2607.21113v1 Announce Type: new Abstract: Autonomous robotic craniotomy requires continuous regulation of tool-tissue interactions to mitigate mechanical overload and thermal damage while mainta

safetyarxiv-cs-ro

Robostral Navigate

arXiv:2607.20785v1 Announce Type: cross Abstract: Deploying navigation systems at scale requires a recipe that minimizes sensor assumptions, generalizes across robot embodiments, and trains efficientl

Robust Asynchronous Q-Learning under Reward and State Corruption via Batching

arXiv:2607.20822v1 Announce Type: new Abstract: Motivated by reinforcement learning in harsh environments, we consider the problem of learning an optimal policy subject to adversarially corrupted feed

Robust Critics: Defending LLMs Against Multi-Turn Attacks

arXiv:2607.20472v1 Announce Type: new Abstract: When a user asks a language model something harmful, is it a genuine attack or a misunderstood but well-meaning question? This ambiguity is one of the c

Routing Subspaces: Auditing Evaluation-to-Deployment Mismatch in Fine-Tuned Language Models

arXiv:2607.20436v1 Announce Type: cross Abstract: Safety evaluations often assume that behavior observed during testing reflects behavior in ordinary use, but fine-tuning can break this assumption. A

Safe and Scalable Multi-Drone Payload Transport via CBF-based Reinforcement Learning with Zero-Shot Sim-to-Real Transfer

arXiv:2607.20665v1 Announce Type: new Abstract: Multi-drone payload transportation has emerged as a promising research paradigm with potential applications in construction, logistics, and disaster res

safetyarxiv-cs-ro

SafeStep: AI-powered Travel Assistance for Elderly People with Frailty or Dementia

arXiv:2607.21156v1 Announce Type: new Abstract: More than a million people in the UK suffer from frailty or dementia, which severely compromise their ability to travel in urban environments. This pape

Safety-oriented sidewalk and road segmentation for smartphone-based assistive navigation

arXiv:2607.21137v1 Announce Type: new Abstract: Independent sidewalk mobility is essential for blind and visually impaired pedestrians (BVIPs), yet smartphone-based assistive navigation requires perce

SAGE: A Socially-Aware Generative Engine for Heterogeneous Multi-Agent Navigation

arXiv:2607.16619v2 Announce Type: replace Abstract: Safe and socially compliant navigation in open human-robot environments requires robots to reason about heterogeneous participants with different dy

safetyarxiv-cs-ro

SalesLoop: Reinforcement Learning from Performance Feedback for Sales Lead Ranking

arXiv:2607.20655v1 Announce Type: cross Abstract: Lead ranking in Customer Relationship Management (CRM) systems faces a persistent challenge: models achieving high offline accuracy often underperform

Scale Up Strategically: Learning Compositional Generalization via Bias-Aware Evaluation and Data Collection for Robotic Manipulation

arXiv:2607.21582v1 Announce Type: cross Abstract: Compositional generalization is essential for robot to follow diverse instructions. However, pretrained policies are known to take shortcuts, deferrin

Self-Supervised Learning of Structured Dynamics from Videos

arXiv:2607.21576v1 Announce Type: new Abstract: Understanding motion in video is a fundamental challenge for visual learning, as frame-to-frame change entangles two sources of dynamics: camera motion

Sidewalk Moments: Are Richer Representations Always More Human-Aligned? Evidence from City-Walk Videos

arXiv:2607.20903v1 Announce Type: new Abstract: We examine whether richer visual representations yield more human-aligned measures of urban engagement, using 61 first-person city-walk videos from YouT

Simple Policy Gradients for Reasoning with Diffusion Language Models

arXiv:2510.04019v3 Announce Type: replace-cross Abstract: Diffusion large language models (dLLMs) represent a promising alternative to autoregressive LLMs; however, the lack of effective post-training

slang.gr as a Large-Scale Crowdsourced Resource for Non-Standard Greek

arXiv:2607.21255v1 Announce Type: cross Abstract: Slang is a central component of everyday language, reflecting linguistic creativity, social identity, and cultural change, yet its dy- namic and non-s

SlerpFlow: Spherical Trajectory Correction for Rectified Flow Inversion

arXiv:2607.21326v1 Announce Type: new Abstract: Rectified-flow-based diffusion transformers, particularly FLUX, have demonstrated outstanding performance in high-quality image generation. However, ach

Spatially Grounded Concept Bottleneck Models for Trustworthy Breast Ultrasound Diagnosis

arXiv:2607.20691v1 Announce Type: cross Abstract: Concept Bottleneck Models provide interpretable-by-design predictions by mediating diagnosis through human-understandable concepts, but in medical ima

SPECTRA: State-Space Exogenous Context and Temporal-Frequency Resolution Architecture for Probabilistic Energy Forecasting

arXiv:2607.20587v1 Announce Type: cross Abstract: Modern power systems increasingly require probabilistic forecasts amid interacting uncertainties from renewable intermittency, flexible demand, market

Statistics of week: “More Americans pay for sports betting apps (5%) than pay for AI… 37% of consumers say none of AI's uses are helpful. … …

Statistics of week: “More Americans pay for sports betting apps (5%) than pay for AI… 37% of consumers say none of AI's uses are helpful. … 2.2% penetration nearly four years after ChatGPT launched. N

Stay hungry, stay foolish. Absolute legend

Stay hungry, stay foolish. Absolute legend For my first post, I’m sharing a letter @NVIDIA signed on why open models matter. AI will transform every industry, power every company, and be built by ever

safetyjim-fan--x

Stokes-Informed Diffusion for Robust Linear Polarization Estimation

arXiv:2607.21239v1 Announce Type: new Abstract: Polarization cues benefit applications such as material detection and de-reflection, yet acquiring them typically requires dedicated hardware. This moti

SuperFlow: Training Flow Matching Models with RL on the Fly

arXiv:2512.17951v3 Announce Type: replace Abstract: Recent progress in flow-based generative models and reinforcement learning (RL) has improved text-image alignment and visual quality. However, curre

Synthetic minority data is redundant or invalid: a data-dependent validity theory and a de-biased test

arXiv:2607.20787v1 Announce Type: cross Abstract: For two decades, the standard remedy for class-imbalanced learning has been to fabricate synthetic minority examples, and the standard evidence of the

thaulab@EEUCA 2026: Who Said What to Whom? A Targeting-Aware Neural-Symbolic Pipeline for Gaming Toxicity Detection

arXiv:2607.20447v1 Announce Type: new Abstract: This paper describes our system for the EEUCA 2026 Shared Task on toxicity classification in gaming chat. We implement a three-stage pipeline combining

safetyarxiv-cs-cl

The Dark Room in the Reward Channel: Dense Prediction Rewards Collapse GRPO-Trained LLM Agents -- and What Actually Works

arXiv:2607.21273v1 Announce Type: new Abstract: Dense per-step supervision is an appealing remedy for sparse-reward, long-horizon LLM agents: reward the agent for predicting its next observation, and

The knowledge that makes AI useful is diffused. It lives with scientists, engineers, clinicians, firms. For AI to benefit from distributed k…

The knowledge that makes AI useful is diffused. It lives with scientists, engineers, clinicians, firms. For AI to benefit from distributed knowledge, it must itself be distributed. Agree with Jensen t

safetyelon-musk--x