Барлық жарияланған талдаулар, егеменді стек зерттеулері, экожүйе жаңалықтары және ЖИ модельдерінің шолулары.
Persistent memory allows an LLM agent to carry experience across conversations, but it also turns a local reasoning mistake into a durable one. During deliberation, an agent may consider a plan, simulate a tool result, report another speaker's belief, and then reject all of…
Humans perceive far more in a scene than what is explicitly depicted: a single glance captures past causes and future trajectories; a quick peek determines if a vehicle can fit between two parked cars; a few seconds of video reveals who holds authority in a room; and a fleeting…
When agents share a reward for completed tasks, reporting unsafe work can reduce the reporter's reward by stopping a task. Audits can make reporting optimal without ensuring that further training teaches a silent team to report. We study this learning problem in a game where any…
An agentic system issues several structurally different kinds of LLM calls. It routes intent, classifies actions, grounds language in a device registry, plans multi-agent pipelines and writes the Python code those pipelines run. The difficulty of these call sites varies by an…
As small language models (SLMs) are increasingly deployed on resource-constrained and on-device platforms, including as components of agentic systems, the integrity of locally stored model parameters becomes an important safety concern. We investigate whether safety-sensitive…
Compositional generalization remains challenging when language models must combine familiar reasoning operations in unfamiliar ways. The Scenario-Based Commonsense Reasoning Evaluation (SCoRE) 2026 tests this ability on three mixed domains absent from training and requires…
We present BEACON-SP, an ontology-grounded Graph Retrieval-Augmented Generation (GraphRAG) framework for clinician-facing decision support in behavioral health settings such as suicide prevention, where effective assessment requires integrating heterogeneous clinical,…
This article develops a failure-mode framework for analyzing how advanced artificial intelligence could contribute to human extinction, irreversible civilizational collapse, or permanent human disempowerment. The central thesis is that catastrophic AI risk does not require…
Agent Skills package procedural guidance and resources for reuse, but a relevant Skill does not necessarily improve task performance. Existing studies characterize Skill content and evaluate downstream performance, yet provide limited explanations of how utility depends on…
Agentic coding makes code generation cheap, but reliable completion remains difficult: the agent that writes the code is a weak judge of whether it is done. We present Humanize, a multi-agent orchestration workflow for agentic coding built around judgement engineering: explicit,…
Multi-agent path planning (MAPP) in continuous environments often relies on roadmaps to balance safety and search efficiency. However, traditional roadmap generation methods, such as lattice grids or standard sampling-based approaches, frequently face a trade-off between graph…
Looped LMs are parameter efficient and promise dynamic computation (saving memory and FLOPs on easy tokens). However, state-of-the-art open Looped LMs trained with this dynamic computation capability (Ouro models) do not realize it in practice as each loop iteration (depth)…
Supervisory governors can interfere with the tool-using agents they regulate. We study this possibility in a controlled file-recovery environment where increases in regulatory intensity trigger experimentally imposed tool failures. A cost-blind governor can turn these failures…
AI agents increasingly operate in environments where they can diagnose failures and improve through experience, yet existing evaluations largely measure what an agent can do at a fixed point in time rather than how effectively it learns. Evaluating self-improvement requires…
Anthropic выкатила Haiku 5.5: модель примерно на 75% дешевле Haiku 4.5 и обходит GPT-6 Luna. Быстрее и умнее Haiku 4.5 Настраиваемая глубина рассуждений Создана для быстрых задач и работы в качестве субагента Для запросов до 100 тысяч токенов цена такая же, как у Luna, но по…
OpenAI disrupted two AI-enabled influence operations that used false-front journalists and a think tank to spread geopolitical messaging.
Привет. Спасибо, что сегодня с нами. Меня зовут Ник. >> А я—Даниэль. >> Мы работаем в команде по взаимодействию с разработчиками здесь , в OpenAI. Позвольте начать с небольшого опроса в зале. Поднимите руку те, у кого начало дня выглядит так: почта пуста, в голове ясность, и вы…
Claude Haiku 5.5-тің Amazon Bedrock-да жаңа мүмкіндіктерімен қолжетімділігі хабарланды
Anthropic анонсировала Claude Haiku 5.5 с 90%-ным снижением цен на токены.
Инженеры из Axiom заставили ИИ управлять Toyota Corolla для обеда в In-N-Out.
OpenAI запускает 'Интеллектуальный интерфейс' в ChatGPT с интерактивными визуалами.
Amazon Quick и Amazon Bedrock реализуют контроль доступа в реальном времени для защиты конфиденциальной информации.
OpenAI is launching a new Intelligent UI feature in ChatGPT that allows the chatbot to answer your questions with interactive visuals. The update, which is rolling out to all users alongside GPT-6, gives ChatGPT the ability to combine a text response with diagrams, charts,…
OpenAI is rolling out GPT-6 with "Intelligent UI," a feature that turns answers into interactive interfaces with charts, buttons, and forms. The model can now respond while still thinking, cutting wait times by 44 percent. Paying customers get GPT-6 Sol, free users get GPT-6…
Anthropic's new Claude Haiku 5.5 crushes its predecessor in benchmarks, jumping from 15.7 to 72.4 percent on the OSWorld computer use test. Token prices drop by up to 90 percent, though a new tokenizer eats into some of those savings by consuming more tokens per task. The…
What’s in a URL anymore, really. For the first time in years, the Internet Corporation for Assigned Names and Numbers - better known as ICANN - is accepting applications for new top-level domains. These are the suffixes at the end of all URLs, and you may know them as things…
⚡ Anthropic представила Claude Haiku 5.5 Claude Haiku 5.5 — самая быстрая и доступная модель в линейке Claude 5.5. Она рассчитана на задачи, где важны скорость, низкая стоимость и масштабирование: массовая обработка запросов, классификация, автоматизация, простые агенты и…
🍌 Google выпустила Nano Banana 2.1 Новая версия стала аккуратнее редактировать отдельные области изображения, лучше сохранять мелкие детали после нескольких изменений и стабильнее работать со сценами, где сразу несколько персонажей. Также улучшили текст и инфографику, добавили…
OpenAI is launching a new user interface that will bring interactive visuals to ChatGPT.