All published intelligence, sovereign stack research, ecosystem news, and AI foundation model reviews.
Good evening again! China is preparing restrictions on the export of its open AI models Today, the Ministry of Commerce and the Ministry of Science and Technology of China published draft amendments to the catalog of technologies prohibited and restricted for export. The decision will be made after November 15th.…
It's really crazy going on in Silicon Valley right now. New personal and assistants appear one after another. Startups are getting billion-dollar valuations, and giants like Meeta and Open AI have entered the real race. [music] But you know what's most interesting? They…
OpenAI is launching the Rosalind Academic Plan, a free offering that's currently open only to research labs at US universities without an enterprise contract. The article OpenAI's new Rosalind Academic Plan gives research labs up to 30 free seats for a year appeared first on The…
Anthropic has given a team from the nonprofit METR inside access to review its AI models and safety practices, the company announced. The article Anthropic gives METR inside access to audit its AI models and safety practices appeared first on The Decoder.
Сегодня на передний план выходят возможности и угрозы, которые несут в себе новые разработки в области искусственного интеллекта. ChatGPT теперь может просматривать вкладки Chrome, что открывает новые
Компания OpenAI обновила расширение ChatGPT для браузера Google Chrome. Чат-бот получил функции полноценного цифрового ассистента и умеет напрямую взаимодействовать с открытыми веб-страницами, избавляя пользователя от рутинных задач. Теперь нейросеть может собирать информацию с…
Wow! It turns out that the GLM-5.3 model sometimes itself, without prompts, begins to reason about how its current actions can become training data and affect its future version/behavior. This was discovered by researchers from Apollo Research, they called it update-aware reasoning. During...
Чат-бот отвечает, агент работает: 8 ИИ для iPhone и Android Агенты берут задачу целиком: от поиска выгодных вариантов до управления экраном. Какой выбрать под вашу рутину, разобрали ниже. ⠀ 🟠Найти и сравнить ⠀ Когда вкладок слишком много, чтобы перебирать их руками. ⠀ Comet —…
Вы дали ИИ-агенту доступ к документам компании. Осталось разобраться, чьи инструкции он выполняет 😏 В сообщении пользователя или документе, который читает модель, могут оказаться вредоносные инструкции. Через них атакующий пытается заставить агента обойти ограничения, выдать…
An AI agent from OpenAI took money from someone else’s wallet and spent it on its own needs. Let us recall that V. Buterin has already offered all cryptocurrency holders a way to save their funds. OpenAI says that the agent worked on a model from the GPT-5.6 Sol family, and was given the task of deploying...
🧠 Meta* выпустила Muse Spark Open – самую мощную open-weight модель из США Модель ориентирована на кодинг, агентские задачи, сложную разработку и длинные цепочки рассуждений. Главные характеристики: • 816 млрд параметров, из них 42 млрд активных; • контекстное окно на 1 млн…
Nvidia has an open world model with 2.6 billion parameters. It lets you turn an image, a text prompt, and a camera motion trajectory into a controllable virtual world. All on a single GPU. - Code available on GitHub. - Research paper published on arXiv. - Suitable for…
🎙 OpenAI has updated Voice - GPT-Live-1.1 has been released. The new version has more voices, improved accents and fewer pauses like “checking...” while answering. GPT-Live-1.1 is already running on ChatGPT, Codex and dots. Also added new voices for: • French; • German; • Hindi; •…
🖥 Tencent rolls out Hy Memory 2.0: open model, framework, and MemHub for long-term AI agent memory. Hunyuan released three things in one release: the Hy-Memory-30B-A3B model, SDK framework, and MemHub application. According to Tencent, this is #1 among open solutions on four...
Ведущий инженер Anthropic, написал руководство о том, как действительно ставить более амбициозные задачи перед ИИ, и поделился готовым промптом. Феликс Ризеберг, работающий над Claude в Anthropic, отмечает, что все советуют быть "амбициознее" при использовании ИИ, но никто не…
Benchmark partners acknowledge that venture capital's traditional rule of not backing rival companies is being tested by the rapid development of AI. Conflicts of interest among portfolio companies are becoming increasingly common as companies quickly...
arXiv:2610.20665v1 Announce Type: new Abstract: Large language model (LLM) agents can solve complex interactive tasks, yet reasoning and tool calls take time. In continuously changing environments, an otherwise correct action may arrive too late. We introduce RealtimeBench, with…
arXiv:2610.20836v1 Announce Type: new Abstract: Evaluator-guided reasoning presents a local control problem at every step: trust the current answer, return a fallback, or spend more inference. We formulate this choice as a compute-aware three-action controller, scoring terminal…
arXiv:2610.20685v1 Announce Type: new Abstract: Symbolic regression searches for an equation that explains observed data. Numerical error is commonly used to guide the search, but it alone can favor an accurate surrogate over a candidate with the right structure. We propose…
arXiv:2610.20676v1 Announce Type: new Abstract: Correctness feedback suppresses the very evidence that specification-completeness auditing is designed to uncover: programs that satisfy a written contract but violate intended behavior. We demonstrate this objective reversal on 12…
arXiv:2610.20780v1 Announce Type: new Abstract: In LLM-based multi-agent systems (MAS), a local mistake can propagate as other agents build on it, turning a correctable error into a failed task. Diagnosing completed runs or repeatedly re-executing them to search for corrections…
arXiv:2610.20926v1 Announce Type: new Abstract: Agent skills turn procedural knowledge into reusable artifacts, allowing LLM agents to accumulate and refine expertise through experience. Building on this, skill evolution iteratively improves skills using execution feedback.…
arXiv:2610.20960v1 Announce Type: new Abstract: Safety in agentic systems is inherently sequential: whether an action is safe depends on prior actions, tool outputs, information disclosures, and interactions among agents. Recent work has shifted agent safety guardrails from…
arXiv:2610.20782v1 Announce Type: new Abstract: The closed world asks AI for specialized competence and lets it manufacture the iid data its theory assumes: a fixed criterion licenses collecting examples until the task is covered. The open world asks for versatility --- taking…
arXiv:2610.20951v1 Announce Type: new Abstract: Decision models answer typed questions about an input with one probability per option, computed in a single forward pass without generating text. Like the fast System 1 of dual-process theory, they answer at once, but their depth…
arXiv:2610.20939v1 Announce Type: new Abstract: Online reinforcement learning (RL) with verifiable rewards has driven rapid progress in mathematics, coding, and tool use, but scaling this success to GUI agents remains challenging due to the high cost of interactive environments…
arXiv:2610.20935v1 Announce Type: new Abstract: Engineering agents must produce artifacts that satisfy exact, coupled constraints. Existing solutions often require agents to make semantic decisions and also derive consequences such as precise geometry or timing from those…
arXiv:2610.20764v1 Announce Type: new Abstract: This paper challenges the understanding capabilities of existing AI paradigms and emerging world models by asking a fundamental question: How quickly can reality change before an intelligent system can no longer reliably perceive,…
arXiv:2610.20681v1 Announce Type: new Abstract: Formal inequality proving requires mathematical reasoning and exact computation, often connected through intermediate lemmas and goal reformulations. We introduce IneqAgent, an agentic framework combining LLM reasoning with…
arXiv:2610.20922v1 Announce Type: new Abstract: Compositional structure is pervasive in nature and across scales: atoms combine into molecules, molecules into cells, and cells into organisms, while further compositional hierarchies appear within each level. More broadly,…