https://ai-news.ypetit.net/sources/ Research publications from the world's largest AI labs and companies, aggregated by AI News. Primary sources only. en Tue, 10 Nov 2026 00:08:45 GMT https://research.google/blog/exploring-the-worlds-languages-with-language-explorer/https://research.google/blog/exploring-the-worlds-languages-with-language-explorer/Mon, 09 Nov 2026 20:29:39 GMTresearchNatural Language Processing https://developer.nvidia.com/blog/make-a-robot-arm-feel-weightless-with-nvidia-holoscan-5/https://developer.nvidia.com/blog/make-a-robot-arm-feel-weightless-with-nvidia-holoscan-5/Mon, 09 Nov 2026 19:09:32 GMTresearchPicture guiding a robot arm by hand while its motors supply just enough torque to support its weight and keep the robot upright. This simple interaction, a... Picture guiding a robot arm by hand while its motors supply just enough torque to support its weight and keep the robo… https://www.anthropic.com/research/provable-inferencehttps://www.anthropic.com/research/provable-inferenceMon, 09 Nov 2026 09:39:55 GMTresearchSharing progress on a powerful type of cryptographic proof known as a zkSNARK, which is uniquely fit for provable inference. http://bair.berkeley.edu/blog/2026/11/09/bair-at-corl-2026/http://bair.berkeley.edu/blog/2026/11/09/bair-at-corl-2026/Mon, 09 Nov 2026 09:00:00 GMTresearchThe Conference on Robot Learning (CoRL) takes place November 9–12, 2026, in Austin, Texas. BAIR faculty, postdocs and students will be presenting research at various sessions, including an Early Career Keynote from Professor Negar Mehr. See below for the full schedule of sessi… https://epoch.ai/publications/inference-performance-indexhttps://epoch.ai/publications/inference-performance-indexMon, 09 Nov 2026 00:00:00 GMTresearchEpoch AI's Inference Performance Index, which goes beyond the underestimates of FLOP/s measures and scores chips on memory and communication https://www.anthropic.com/research/teaching-ai-to-build-scientific-modelshttps://www.anthropic.com/research/teaching-ai-to-build-scientific-modelsSat, 07 Nov 2026 23:12:43 GMTresearchAI for Science grantees Jakob Macke, Daniel Gedon, and Stefan Wahl on how AI is starting to change how scientific models are built and interpreted. https://developer.nvidia.com/blog/speed-up-image-generation-with-flux-3-image-klein-on-nvidia-blackwell-gpus/https://developer.nvidia.com/blog/speed-up-image-generation-with-flux-3-image-klein-on-nvidia-blackwell-gpus/Fri, 06 Nov 2026 22:20:20 GMTresearchWhen deploying diffusion models in design tools, image editors, and visual production workflows, developers must account for inference latency. Generating an... When deploying diffusion models in design tools, image editors, and visual production workflows, developers must acc… https://www.anthropic.com/research/components-of-agent-memoryhttps://www.anthropic.com/research/components-of-agent-memoryFri, 06 Nov 2026 21:15:32 GMTresearchFor Claude to learn from experience, it has to decide what’s worth remembering before it knows what’s next. Here’s how “dreaming” helps it organize memory and correct mistakes. https://developer.nvidia.com/blog/accelerating-reinforcement-learning-weight-refit-with-the-reshard-collective/https://developer.nvidia.com/blog/accelerating-reinforcement-learning-weight-refit-with-the-reshard-collective/Fri, 06 Nov 2026 18:03:26 GMTresearchDuring reinforcement learning (RL) post-training, training trajectories are generated on the fly by running inference with the current policy model. For... During reinforcement learning (RL) post-training, training trajectories are generated on the fly by running inference wit… https://www.anthropic.com/research/project-eve-claude-plays-mmorpgshttps://www.anthropic.com/research/project-eve-claude-plays-mmorpgsFri, 06 Nov 2026 17:09:29 GMTresearchWe put Claude in two EVE games to study how AI agents handle markets and teamwork, where they break down, and what safeguards they might need. https://www.anthropic.com/research/how-researchers-use-claudehttps://www.anthropic.com/research/how-researchers-use-claudeFri, 06 Nov 2026 15:55:34 GMTresearchWe analyzed 207,000 research sessions to learn how Claude is used in science, why some tasks get more use than others, and how AI could speed up discovery. https://www.anthropic.com/research/accelerating-scientific-researchhttps://www.anthropic.com/research/accelerating-scientific-researchFri, 06 Nov 2026 10:23:12 GMTresearchLabs at Stanford and MIT built Claude-powered systems that run genome-wide studies in minutes and automate gene cluster interpretation. https://allenai.org/blog/toyblockshttps://allenai.org/blog/toyblocksFri, 06 Nov 2026 08:00:00 GMTresearchToyBlocks is an open simulation framework for studying robot learning through automated task generation, training demonstrations, and evaluation in a controlled tabletop world. https://www.anthropic.com/research/investigating-vault-particles-with-claudehttps://www.anthropic.com/research/investigating-vault-particles-with-claudeFri, 06 Nov 2026 01:53:44 GMTresearchWe’re sharing another finding from Anthropic’s biology research lab: the first 3D structure of a vault particle found in prokaryotic cells. https://www.anthropic.com/research/early-progress-drug-developmenthttps://www.anthropic.com/research/early-progress-drug-developmentFri, 06 Nov 2026 00:33:28 GMTresearchOver the past several months, we’ve partnered with six organizations to accelerate curing disease. We’ve also started drug discovery efforts of our own, including programs for rare childhood forms of muscular dystrophy. In this post, we provide an update on our progress. https://epoch.ai/publications/huawei-horinger-china-compute-gaphttps://epoch.ai/publications/huawei-horinger-china-compute-gapFri, 06 Nov 2026 00:00:00 GMTresearchHuawei’s AI data center in Horinger, Inner Mongolia, uses Ascend chips and has an estimated AI computing capacity that is just 7% that of the world’s largest AI data center, SpaceX’s Colossus 2 in Tennessee. https://deepmind.google/blog/gemini-omni-12-flash-fast-and-cost-effective-with-longer-clips/https://deepmind.google/blog/gemini-omni-12-flash-fast-and-cost-effective-with-longer-clips/Thu, 05 Nov 2026 17:05:16 GMTresearch https://developer.nvidia.com/blog/configure-certify-and-operate-kubernetes-gpu-clusters-with-nvidia-dsx-os/https://developer.nvidia.com/blog/configure-certify-and-operate-kubernetes-gpu-clusters-with-nvidia-dsx-os/Thu, 05 Nov 2026 17:00:00 GMTresearchOne underperforming GPU can slow an entire distributed training job. The node may pass basic health checks and show normal utilization, leaving operators to... One underperforming GPU can slow an entire distributed training job. The node may pass basic health checks and show n… https://research.google/blog/detecting-signs-of-hypertension-using-wearable-sensors/https://research.google/blog/detecting-signs-of-hypertension-using-wearable-sensors/Thu, 05 Nov 2026 16:01:08 GMTresearchHealth & Bioscience https://www.microsoft.com/en-us/research/blog/formally-verifying-agent-written-full-stack-web-apps-with-pdd-kit/https://www.microsoft.com/en-us/research/?p=1191852Thu, 05 Nov 2026 16:00:00 GMTresearchAI agents can generate code at a scale that’s hard for humans to review line-by-line. PDD Kit helps ensure the code is doing what the developer intended by creating formal specifications and testing against them. It converts developer intent into a Verus specification, proves … https://developer.nvidia.com/blog/five-ways-nvidia-co-design-maximizes-the-value-of-every-byte-in-the-ai-factory/https://developer.nvidia.com/blog/five-ways-nvidia-co-design-maximizes-the-value-of-every-byte-in-the-ai-factory/Thu, 05 Nov 2026 16:00:00 GMTresearchIn an AI factory, high bandwidth memory (HBM)is the fastest, most expensive, and most contested memory tier. Every byte spent on a redundant weight copy, an... In an AI factory, high bandwidth memory (HBM)is the fastest, most expensive, and most contested memory tier. Every by… https://www.amazon.science/blog/hamed-karimi-wins-ecml-pkdd-test-of-time-awardhttps://www.amazon.science/blog/hamed-karimi-wins-ecml-pkdd-test-of-time-awardThu, 05 Nov 2026 14:57:23 GMTresearchAmazon senior applied scientist and coauthors showed that an overlooked mathematical condition from 1963 proves why frequently used machine-learning optimization methods work well. https://www.anthropic.com/research/alignment-assessment-ukaisihttps://www.anthropic.com/research/alignment-assessment-ukaisiThu, 05 Nov 2026 14:10:48 GMTresearchIn a UK AISI evaluation, Mythos 5 took unsanctioned actions that impacted real people and organizations. In this post, we investigate these behaviors from Claude Mythos 5 further, to understand what they reveal about the model’s alignment. https://machinelearning.apple.com/research/faster-retrievalfaster-retrievalThu, 05 Nov 2026 00:00:00 GMTresearchMaximum inner product search (MIPS) sits at the core of recommendation systems, search, and retrieval-augmented generation (RAG). Given a query vector, MIPS systems search within a database of millions of candidate key vectors for the one that best matches the query. The origi… https://developer.nvidia.com/blog/how-nvidia-uses-ai-agents-to-streamline-ai-factory-operations/https://developer.nvidia.com/blog/how-nvidia-uses-ai-agents-to-streamline-ai-factory-operations/Wed, 04 Nov 2026 22:00:17 GMTresearchAs GPU clusters scale, operations span compute, network, hardware, and scheduling. Site reliability engineers (SREs) often have to examine multiple dashboards... As GPU clusters scale, operations span compute, network, hardware, and scheduling. Site reliability engineers (SREs… https://developer.nvidia.com/blog/nvidia-cuda-13-5-brings-nvidia-vera-rubin-ga-support/https://developer.nvidia.com/blog/nvidia-cuda-13-5-brings-nvidia-vera-rubin-ga-support/Wed, 04 Nov 2026 18:46:07 GMTresearchAs artificial intelligence and accelerated computing continue to push the boundaries of what's possible, the latest NVIDIA CUDA release arrives to unlock... As artificial intelligence and accelerated computing continue to push the boundaries of what’s possible, the latest NVID… https://research.google/blog/agile-deliberation-how-ai-helps-humans-define-the-gray-areas/https://research.google/blog/agile-deliberation-how-ai-helps-humans-define-the-gray-areas/Wed, 04 Nov 2026 18:07:05 GMTresearchData Mining & Modeling https://developer.nvidia.com/blog/how-nvidia-factory-operations-blueprint-coordinates-ai-agents-to-investigate-production-issues/https://developer.nvidia.com/blog/how-nvidia-factory-operations-blueprint-coordinates-ai-agents-to-investigate-production-issues/Wed, 04 Nov 2026 17:00:00 GMTresearchA Siemens report estimates that unplanned downtime costs an average large plant in the surveyed sectors $253 million per year. Whether investigating a failed... A Siemens report estimates that unplanned downtime costs an average large plant in the surveyed sectors $253 million… https://developer.nvidia.com/blog/use-nvidia-nemo-to-turn-agent-failures-into-tests-and-harness-fixes/https://developer.nvidia.com/blog/use-nvidia-nemo-to-turn-agent-failures-into-tests-and-harness-fixes/Wed, 04 Nov 2026 16:00:00 GMTresearchTurn production evidence into prioritized issues and custom evaluations that help you improve prompts, tools, models, memory, skills, and harness code. When you... Turn production evidence into prioritized issues and custom evaluations that help you improve prompts, tools, mod… https://allenai.org/blog/google-climate-partnershiphttps://allenai.org/blog/google-climate-partnershipWed, 04 Nov 2026 08:00:00 GMTresearchWith funding from Google.org and technical support from Google Research, Ai2 is developing NeuralGCM+, an open-source model designed to combine physics and AI to simulate the atmosphere and ocean together for faster, more reliable long-term climate projections. https://epoch.ai/data-insights/tpu-production-doublinghttps://epoch.ai/data-insights/tpu-production-doublingWed, 04 Nov 2026 00:00:00 GMTresearchBroadcom’s shipments of TPU by unit volume increased by about 65% quarter over quarter to roughly 800,000 chips, while compute output doubled, driven by a shift to more powerful chips. https://epoch.ai/publications/eci-bayesian-methodologyhttps://epoch.ai/publications/eci-bayesian-methodologyWed, 04 Nov 2026 00:00:00 GMTresearchA Bayesian model gives the ECI better-justified uncertainty estimates and lets us compare models directly. https://www.microsoft.com/en-us/research/podcast/learn-the-rules-the-hacker-mindset/https://www.microsoft.com/en-us/research/?p=1192298Tue, 03 Nov 2026 17:14:38 GMTresearchResearch Manager Mike Walker reflects on a career spanning hacker culture, cybersecurity, and AI, sharing lessons about trust, teamwork, leadership, and why solving difficult problems requires curiosity, judgment, agility—and an understanding of the rules. The post Learn the r… https://deepmind.google/blog/agentic-cheminformatics-with-google-antigravity/https://deepmind.google/blog/agentic-cheminformatics-with-google-antigravity/Tue, 03 Nov 2026 17:01:49 GMTresearchLearn how Google Antigravity and our Cheminformatics Skills accelerate drug discovery workflows, search compounds, and propose novel rigidified macrocycles. https://allenai.org/blog/statebenchhttps://allenai.org/blog/statebenchTue, 03 Nov 2026 08:00:00 GMTresearchStateBench tests how language models track changing information across long documents at four difficulty levels, revealing where models struggle and how targeted training helps recurrent and hybrid architectures handle longer sequences. https://developer.nvidia.com/blog/diagnose-distributed-ai-workload-failures-with-nvidia-attribution/https://developer.nvidia.com/blog/diagnose-distributed-ai-workload-failures-with-nvidia-attribution/Mon, 02 Nov 2026 23:39:34 GMTresearchIn distributed AI workloads, a single error can produce a cascade of symptoms. A GPU or fabric event may surface as NCCL errors on other ranks, scheduler state... In distributed AI workloads, a single error can produce a cascade of symptoms. A GPU or fabric event may surface a… https://developer.nvidia.com/blog/keep-distributed-training-running-with-automated-workload-recovery/https://developer.nvidia.com/blog/keep-distributed-training-running-with-automated-workload-recovery/Mon, 02 Nov 2026 22:43:05 GMTresearchAs distributed AI training scales, failures across hardware and software dependencies increasingly disrupt training progress, making fast, reliable recovery... As distributed AI training scales, failures across hardware and software dependencies increasingly disrupt training p… https://developer.nvidia.com/blog/making-kubernetes-ai-ready-for-nvidia-vera-rubin-on-day-one/https://developer.nvidia.com/blog/making-kubernetes-ai-ready-for-nvidia-vera-rubin-on-day-one/Mon, 02 Nov 2026 18:00:00 GMTresearchConfiguring and optimizing Kubernetes for new GPU hardware such as NVIDIA Vera Rubin is complex and error-prone. Drivers, kernels, runtimes, networking,... Configuring and optimizing Kubernetes for new GPU hardware such as NVIDIA Vera Rubin is complex and error-prone. Drivers,… https://developer.nvidia.com/blog/build-world-models-for-interactive-simulation-with-nvidia-cosmos/https://developer.nvidia.com/blog/build-world-models-for-interactive-simulation-with-nvidia-cosmos/Mon, 02 Nov 2026 17:00:00 GMTresearchPhysical AI developers working with world models face two distinct challenges: adapting models to their sensors, tasks, and data, and getting those models into... Physical AI developers working with world models face two distinct challenges: adapting models to their sensors, t… https://developer.nvidia.com/blog/support-new-ai-frameworks-faster-using-nvidia-workload-map/https://developer.nvidia.com/blog/support-new-ai-frameworks-faster-using-nvidia-workload-map/Mon, 02 Nov 2026 16:00:00 GMTresearchNVIDIA DSX AI Factory Platform brings together technologies to help the ecosystem design, build, and operate more efficient AI factories, including open source,... NVIDIA DSX AI Factory Platform brings together technologies to help the ecosystem design, build, and operate more… https://www.anthropic.com/research/vibe-physicshttps://www.anthropic.com/research/vibe-physicsMon, 02 Nov 2026 10:12:52 GMTresearchA Harvard physicist guided Claude Opus 4.5 through a full quantum chromodynamics research project in two weeks. https://www.anthropic.com/research/claude-shaped-sciencehttps://www.anthropic.com/research/claude-shaped-scienceMon, 02 Nov 2026 10:11:59 GMTresearchMatthew Schwartz describes what happened when he stopped fighting Claude and allowed Claude to find “Claude-shaped” problems. This led him to build BootLoops, a toolkit for exact calculations in quantitative science, which he has been applying across scientific fields alongsid… https://www.anthropic.com/research/Claude-accelerates-protein-designhttps://www.anthropic.com/research/Claude-accelerates-protein-designMon, 02 Nov 2026 10:11:06 GMTresearchTwo results show how Claude helps life sciences researchers speed up their work: designing protein binders from scratch and analyzing NMR and LC-MS data to speed chemical analysis. https://allenai.org/blog/nextcloudhttps://allenai.org/blog/nextcloudMon, 02 Nov 2026 08:00:00 GMTresearchNextcloud chose Olmo 3 for fully open, self-hosted AI, earning its highest Ethical AI rating while making local AI practical on consumer hardware. https://arxiv.org/abs/2611.02102v3https://arxiv.org/abs/2611.02102v3Sun, 01 Nov 2026 18:16:47 GMTresearchScaling computer-use agents from individual actions to complex software work requires experience that connects decisions to complete outcomes. We present NorthCUA, an integrated system for turbocharging computer-use agents through experience at scale. Its Experience Flywheel s… https://arxiv.org/abs/2611.01112v1https://arxiv.org/abs/2611.01112v1Sat, 31 Oct 2026 16:59:05 GMTresearchVision-language models can interpret instructions and plan manipulation tasks, yet struggle when acting directly as robot policies. On RoboDojo, a large planning-execution gap separates high-level task decisions from physical execution: removing arm execution raises GPT-6 Astr… https://arxiv.org/abs/2611.00562v1https://arxiv.org/abs/2611.00562v1Sat, 31 Oct 2026 02:37:09 GMTresearchTabular foundation models (TFMs) enable in-context prediction on unseen tabular datasets without task-specific training. Mechanistic work has shown that increasing the depth of the ICL Transformer yields weak gains, leaving the main headroom for gains in the synthetic pretrain… https://developer.nvidia.com/blog/build-a-validated-kubernetes-runtime-for-nvidia-dynamo/https://developer.nvidia.com/blog/build-a-validated-kubernetes-runtime-for-nvidia-dynamo/Fri, 30 Oct 2026 21:35:31 GMTresearchTeams building AI applications need a repeatable path from a model configuration to a working inference endpoint. NVIDIA Dynamo is a distributed inference... Teams building AI applications need a repeatable path from a model configuration to a working inference endpoint. NVIDI… https://developer.nvidia.com/blog/run-ai-models-for-edge-application-development-with-just-a-few-commands/https://developer.nvidia.com/blog/run-ai-models-for-edge-application-development-with-just-a-few-commands/Fri, 30 Oct 2026 20:33:18 GMTresearchAI is expanding beyond the data center as developers build a new generation of agents to act at the edge: monitoring driver attention and cabin conditions,... AI is expanding beyond the data center as developers build a new generation of agents to act at the edge: monitoring d… https://developer.nvidia.com/blog/build-specialized-agents-on-a-stack-you-control-with-nvidia-nemoclaw/https://developer.nvidia.com/blog/build-specialized-agents-on-a-stack-you-control-with-nvidia-nemoclaw/Fri, 30 Oct 2026 16:00:00 GMTresearchBuilding an autonomous agent requires a model, a harness, and domain-specific tools. To deploy an agent more securely, developers need a runtime that enforces... Building an autonomous agent requires a model, a harness, and domain-specific tools. To deploy an agent more secure…