All entries

No Safe Dose: How Training Data Drives Unsafe Image Generation

arXiv:2605.28137v1 Announce Type: new Abstract: Text-to-image models trained on large-scale data often inevitably ingest unsafe content. While some people observe input-output amplifications, it remai

safetyarxiv-cs-cv

Noise Scheduling as Information-Guided Allocation in Diffusion Training

DGX agent

arXiv:2602.18647v2 Announce Type: replace-cross Abstract: We introduce InfoNoise, an online adaptive noise schedule for diffusion training that reallocates optimization effort toward noise levels wher

researcharxiv-cs-ai

28 May 2026

Not All NVFP4 QAT Recipes Are Equal: How Architecture and Scale Shape Model Quality for Anomaly Segmentation

arXiv:2605.27616v1 Announce Type: cross Abstract: Real-time anomaly segmentation demands both high recall and efficient low-precision inference. We study the three-way interaction of model architectur

Not All Pixels Are Equal: Pixel-wise Meta-Learning for Medical Segmentation with Noisy Labels

arXiv:2511.18894v5 Announce Type: replace-cross Abstract: Medical image segmentation is crucial for clinical applications, but it is frequently disrupted by noisy annotations and ambiguous anatomical

Note, I am posted this to encourage those who enjoy getting incredible performance out of hardware to join SpaceX

Elon Musk posted on X to recruit engineers and technical talent interested in optimizing hardware performance to join SpaceX. The post encourages those with expertise in achieving exceptional performa

industryelon-musk--x

NUCLEUS-MoE: Unified Model of Pool Boiling for Liquid Cooling

arXiv:2605.27722v1 Announce Type: new Abstract: Two-phase boiling enables heat transfer rates an order of magnitude higher than single-phase cooling, but it remains difficult to model due to the stron

researcharxiv-cs-lg

Nutanix’s stock inches up on solid earnings and revenue beat and growing customer acquisitions

Hyperconverged cloud infrastructure company Nutanix Inc. delivered a mostly upbeat earnings report today, easily beating estimates on profit and revenue, and investors liked the news as the company’s

industrysiliconangle

Hardware

NVIDIA Research Advances Robotics From Simulation to the Real World

Robotics is entering a new phase: moving from controlled demos and scripted automation toward generalizable, reliable embodied autonomy in the real world. At the International Conference on Robotics a

hardwarenvidia-blog

Object-Centric Vision Token Pruning for Vision Language Models

arXiv:2511.20439v2 Announce Type: replace-cross Abstract: In Vision Language Models (VLMs), vision tokens are quantity-heavy yet information-dispersed compared with language tokens, thus consume too m

Applications

ObjFiller3D: Scaling 3D Object Inpainting to Dense Multi-View Consistency

arXiv:2508.18271v2 Announce Type: replace Abstract: 3D object inpainting is commonly achieved via multi-view 2D image completion, yet independently inpainted views often suffer from cross-view inconsi

applicationsarxiv-cs-cv

OccuReward: LLM-Guided Occupant-Centric Reward Shaping for Demographic Equity in Grid-Interactive Buildings

arXiv:2605.28168v1 Announce Type: new Abstract: Large language models (LLMs) have demonstrated promising capability in generating reward functions for deep reinforcement learning (DRL)-based building

Ocean4Rec: Offline LLM-Derived OCEAN Profiles for Request-Time VOD Reranking

arXiv:2605.27429v1 Announce Type: cross Abstract: Industrial video-on-demand (VOD) recommenders need richer content understanding, but LLM-as-reranker designs repeat prompt construction, token generat

Off-Policy Learning to Reason Works Because It Is More Pessimistic Than You Think

arXiv:2605.28150v1 Announce Type: new Abstract: Large scale reinforcement learning has become a central tool for improving reasoning in large language models. At this scale, generation is often lagged

safetyarxiv-cs-lg

Hardware

Official @NVIDIAAI GLM5.1-NVFP4 spotted on @huggingface 🤩 https://huggingface.co/nvidia/GLM-5.1-NVFP4

NVIDIA has released GLM-5.1-NVFP4, a quantized version of the GLM-5.1 model, now available on Hugging Face. The model appears to use NVFP4 (NVIDIA's floating-point 4-bit) quantization format, designed

hardwareclem-delangue--x

OGER: A Robust Offline-Guided Exploration Reward for Hybrid Reinforcement Learning

arXiv:2604.18530v2 Announce Type: replace Abstract: Recent advancements in Reinforcement Learning with Verifiable Rewards (RLVR) have significantly improved Large Language Model (LLM) reasoning, yet m

safetyarxiv-cs-ai

okay i think this is a much better visualization of what i mean by 'log-centric agent architecture'

okay i think this is a much better visualization of what i mean by 'log-centric agent architecture' babyagi has ~200 citations, but 0 papers... i just published my first paper on arXiv 😆 'The Log is t

agentsyohei-nakajima--x

Okta reports Q1 revenue up 11% YoY to 765M, vs. 752M est., says the agentic AI build-out is spiking demand for its identity tools; OKTA jumps 8%+ after hours (Samantha Subin/CNBC)

Samantha Subin / CNBC: Okta reports Q1 revenue up 11% YoY to 765M, vs. 752M est., says the agentic AI build-out is spiking demand for its identity tools; OKTA jumps 8%+ after hours — Okta beat Wall St

agentstechmeme

Okta tops estimates and raises outlook as SentinelOne slides on weak guidance

Shares of Okta Inc. rose more than 6% in late trading today, while SentinelOne Inc. fell more than 18%, as the two cybersecurity companies reported quarterly results hours apart with sharply different

industrysiliconangle

{Omega}-QVLA: Robust Quantization for Vision-Language-Action Models via Composite Rotation and Per-step Scaling

model-releasesarxiv-cs-cv

arXiv:2605.28803v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models unify perception, reasoning, and control within a single policy, yet their multi-billion-parameter backbones and dif

OmniEgo-R^2: A Routed Reasoning Framework for the 1st Cross-Domain EgoCross Challenge at CVPR 2026

arXiv:2605.24481v2 Announce Type: replace Abstract: The 1st Cross-Domain EgoCross Challenge at EgoVis, CVPR 2026 evaluates whether multimodal large language models can reason over egocentric videos ac

researcharxiv-cs-cv

Local Ai

OmniVerifier-M1: Multimodal Meta-Verifier with Explicit Structured Recalibration

arXiv:2605.28805v1 Announce Type: cross Abstract: Visual outcomes are increasingly central to multimodal large language models, making reliable and fine-grained verification essential for scaling gene

local-aiarxiv-cs-ai

On Compositional Learning Behaviours in Formal Mathematics

model-releasesarxiv-cs-cl

arXiv:2605.28512v1 Announce Type: new Abstract: Self-evolving scientific agents capable of conquering the hard tail of formal mathematics require Compositional Learning Behaviours (CLBs) -- the capaci

On the Equivariant Learning of the Q-tensor Order Parameter

model-releasesarxiv-cs-cv

arXiv:2605.27679v1 Announce Type: cross Abstract: We construct and evaluate group-equivariant neural networks for the prediction of the two-dimensional Q-tensor order parameter of nematic liquid cryst

On the Fallacy of Global Token Perplexity in Spoken Language Model Evaluation

arXiv:2601.06329v2 Announce Type: replace-cross Abstract: Generative spoken language models pretrained on large-scale raw audio can continue a speech prompt with appropriate content while preserving a

On the Intrinsic Limits of Transformer Image Embeddings in Non-Solvable Spatial Reasoning

arXiv:2601.03048v2 Announce Type: replace-cross Abstract: Vision Transformers (ViTs) excel in semantic recognition but exhibit systematic failures in spatial reasoning tasks such as mental rotation. W

On the Learnability of Test-Time Adaptation: A Recovery Complexity Perspective

arXiv:2605.28057v1 Announce Type: cross Abstract: Test-time adaptation (TTA) aims to adapt models to maintain reliable performance on non-stationary test streams without requiring labeled data. Despit

On the Origin of Synthetic Information by Means of Steganographic Inheritance

arXiv:2605.27551v1 Announce Type: new Abstract: The origin of species has been the mystery of mysteries in natural science. By analogy, the origin of synthetic information, we suggest, is the mystery

On the Subgaussianity of Quantized Linear Maps: An AI-Assisted Note

arXiv:2605.27563v1 Announce Type: cross Abstract: This short note presents a dimension-independent subgaussian concentration bound for Gaussian vectors under coordinate-wise nonlinear mappings. Discov

One fascinating thing we saw that doesn't seem to be (?) in any literature is the effect of Whitehead's second lemma on de Sitter deformatio…

One fascinating thing we saw that doesn't seem to be (?) in any literature is the effect of Whitehead's second lemma on de Sitter deformation: https://en.wikipedia.org/wiki/Whitehead%27s_lemma_(Lie_al

industryemad-mostaque--x

One of the best ways to learn what LangSmith Engine is capable of is to talk to the team that built it. Join @bentannyhill for a live sessio…

One of the best ways to learn what LangSmith Engine is capable of is to talk to the team that built it. Join @bentannyhill for a live session on June 11th and see how your team can automate the agent

agentsharrison-chase--x

Applications

Online Irregular Multivariate Time Series Forecasting via Uncertainty-Driven Dual-Expert Calibration

arXiv:2605.28603v1 Announce Type: cross Abstract: Irregular multivariate time series forecasting is critical in many real-world applications, where time series are irregularly sampled and exhibit dyna

applicationsarxiv-cs-ai

@OpenAI Foundation just launched with a 25B commitment and equity in OpenAI valued at ~130B. Sounds like a philanthropy giant. But look cl…

@OpenAI Foundation just launched with a 25B commitment and equity in OpenAI valued at ~130B. Sounds like a philanthropy giant. But look closer: the big money goes to programs they control internally (

safetygary-marcus--x

OpenAI’s Frontier Governance Framework

OpenAI's Frontier Governance Framework outlines the organization's approach to managing risks associated with advanced AI systems, including safety, security, and responsible deployment practices. The

safetyopenai

OpenCode Now Supports DigitalOcean Inference Router for Intelligent Model Routing

OpenCode now integrates with DigitalOcean's Inference Router, enabling intelligent routing of AI model requests across distributed infrastructure. This integration allows developers to optimize model

industrydigitalocean

Hardware

OpenURMA: A Clean-Room Open Implementation of the Unified Bus Protocol

arXiv:2605.28717v1 Announce Type: new Abstract: Modern datacenter RDMA is bottlenecked at the network interface, not the wire. A NIC running RoCE or InfiniBand holds per-connection state for every (ap

hardwarearxiv-cs-ai

Operational AI Deployment Assurance: Governance-State Orchestration Under Threshold-Sensitive Deployment Conditions -- A Governance Framework for High-Stakes AI Systems

arXiv:2605.27827v1 Announce Type: new Abstract: AI governance frameworks increasingly emphasize fairness, transparency, accountability, and lifecycle risk management in high-stakes domains. However, m

safetyarxiv-cs-ai

Applications

OphIn-500K: Curating Web-Scale Visual Instructions for Scaling Ophthalmic Multimodal Large Language Models

arXiv:2605.27916v1 Announce Type: cross Abstract: The advancement of general medical Multimodal Large Language Models (MLLMs) has shown great potential for building conversational assistants to suppor

applicationsarxiv-cs-cl

Optimal and Diffusion Transports in Machine Learning

arXiv:2512.06797v2 Announce Type: replace-cross Abstract: Several problems in machine learning are naturally expressed as the design and analysis of time-evolving probability distributions. This inclu

Optimal Data Acquisition for Reinforcement Learning: A Large Deviations Perspective

arXiv:2605.28675v1 Announce Type: new Abstract: Data acquisition efficiency is a central challenge in deploying reinforcement learning in business and healthcare operations, where interactions are cos

safetyarxiv-cs-lg

Optimal LTLf Synthesis

arXiv:2605.11544v2 Announce Type: replace Abstract: Strategy synthesis typically follows an all-or-nothing paradigm, returning unrealisable whenever a specification cannot be guaranteed in an uncertai

Optimal ridge regularization revisited

model-releasesarxiv-cs-lg

arXiv:2605.28679v1 Announce Type: new Abstract: We consider L^2-regularized linear (ridge) regression over a finite data sample X with bounded covariance and linear prediction targets y with additive

Opus 4.8 formulated the hypotheses in advance, conducting data cleaning, did research on references, conducted analyses, did robustness chec…

model-releasesethan-mollick--x

Opus 4.8 formulated the hypotheses in advance, conducting data cleaning, did research on references, conducted analyses, did robustness checks, and put out the whole paper in LaTEX style. GPT-5.5 foun

Opus 4.8 is now supported in Hermes Agent ^_^

Nous Research has announced support for Opus 4.8 in the Hermes Agent framework. This update enables the Hermes Agent to utilize Anthropic's Opus 4.8 model, expanding the available model options for us

agentsnous-research--x

Tools

Opus 4.8 on AI Gateway

Vercel announced support for Claude Opus 4.8 on its AI Gateway, enabling developers to integrate Anthropic's latest large language model through Vercel's unified API platform. This update allows seaml

toolsvercel-blog

OR-Space: A Full-Lifecycle Workspace Benchmark for Industrial Optimization Agents

arXiv:2605.28158v1 Announce Type: new Abstract: Large language model (LLM) agents are increasingly used to assist with operations research (OR) modeling, yet existing OR-oriented benchmarks often redu

OralAgent: Integrating Reasoning, Tools, and Knowledge for Interactive Dental Image Analysis

model-releasesarxiv-cs-cl

arXiv:2605.27378v1 Announce Type: new Abstract: Dental image analysis plays a pivotal role in supporting accurate diagnosis and treatment planning in oral healthcare. Although recent advances have pro

Orchid Security targets AI agent sprawl with new identity governance tools

Orchid Security Inc. today extended its Identity Control Plane with a set of capabilities aimed at governing artificial intelligence agents, saying existing identity and access management models canno

agentssiliconangle

Local Ai

OSP-Next: Efficient High-Quality Video Generation with Sparse Sequence Parallelism, HiF8 Quantization, and Reinforcement Learning

arXiv:2605.28691v1 Announce Type: new Abstract: Diffusion Transformers achieve strong video generation quality, but the quadratic cost of full attention limits efficiency. We introduce OSP-Next, an ef

local-aiarxiv-cs-cv