Барлық жарияланған талдаулар, егеменді стек зерттеулері, экожүйе жаңалықтары және ЖИ модельдерінің шолулары.
arXiv:2609.28654v1 Announce Type: new Abstract: Object permanence and solidity are hallmarks of human cognitive priors. Recent studies show that video generation models, a paradigmatic class of current world models, have begun to show emerged reasoning abilities, making them…
arXiv:2609.28921v1 Announce Type: new Abstract: Protein modification requires navigating an immense sequence space, yet wet-lab validation remains low-throughput and costly. Although computational paradigms including protein language models (PLMs), large language models (LLMs),…
arXiv:2609.28771v1 Announce Type: new Abstract: Large language models (LLMs) have demonstrated strong capabilities in financial analysis and reasoning, inspiring recent advances in agent-based trading frameworks. While these systems show promise, prior approaches either…
arXiv:2609.28609v1 Announce Type: new Abstract: Role-playing agents based on large language models have been widely applied in areas such as personalized assistance and social simulation. Recent RL methods typically train on a fixed scenario pool collected before learning…
arXiv:2609.28547v1 Announce Type: new Abstract: Policy interventions propagate through public communication, institutional decisions, and stakeholder responses, yet datasets for financial multi-agent simulation rarely connect these processes to temporally aligned historical…
arXiv:2609.28942v1 Announce Type: new Abstract: Personalized value alignment has become increasingly important as large language models (LLMs) are expected to accommodate diverse user preferences. However, existing methods typically align model outputs with a static value…
arXiv:2609.28570v1 Announce Type: new Abstract: Reinforcement learning (RL) is widely used to sharpen reasoning in multimodal large language models (MLLMs), yet its effect on hallucination is uneven. We trace this to two weak points in the \emph{correction chain} from reward to…
arXiv:2609.28575v1 Announce Type: new Abstract: Long-conversation memory benchmarks increasingly test recall and prompted knowledge updates, and recent work studies evolving user beliefs and memory state. TWIST is a proposed benchmark suite for a complementary, unmeasured…
arXiv:2609.28919v1 Announce Type: new Abstract: Harnesses, the products that run AI coding agents, are multiplying, and enterprises are rolling them out to their employees: what started as pilots with a few hundred seats is scaling to tens of thousands. Most enterprises do not…
arXiv:2609.29007v1 Announce Type: new Abstract: Tool-using agents are continually updated with new interaction data. After each policy update, however, previously estimated action credits may become stale. Recomputing them from scratch can require many additional tool calls and…
arXiv:2609.29015v1 Announce Type: new Abstract: Decentralized LLM-based multi-agent systems coordinate through local interactions, but an agent can remain responsive while its task-solving quality persistently degrades. Such gray failures require protecting current tasks before…
arXiv:2609.28557v1 Announce Type: new Abstract: DNA sequencing pipelines, spanning quality control, alignment, variant calling, and annotation, are now reliably executed by workflow management systems that orchestrate established bioinformatics tools at scale. What remains…
Wayfair uses OpenAI models to improve ecommerce support and product catalog accuracy, automating ticket triage and enhancing millions of product attributes at scale.
AI-режиссер видео обеспечивает визуальную непрерывность в длинных нарративах.
BottleCap AI выпустила ThinkingCap-Qwen3.8-27B с меньшими затратами токенов, но с незначительным снижением точности.
PrismML разработала компактные языковые модели для смарт-очков Qualcomm.
Prism's larger goal is open-weight AI that runs on devices and makes better use of the computing power they already have.
WhisperX от AWS предлагает улучшенную транскрипцию речи с метками спикеров и точными временными метками.
The notice would allow Oracle to delay payments should the facility miss its 2028 target to come online.
ИИ начали проверять на реалистичных разговорах о психическом здоровье — появился новый открытый бенчмарк MentalHea....
😬 Даже люди, которые пользуются ИИ каждый день, всё равно его опасаются — особенно в США Новое исследование Gallup показало необычный парадокс: 68% американцев, ежедневно использующих ИИ, испытывают беспокойство из-за технологии. Среди всех опрошенных американцев этот…
TechCrunch Founder Summit is a full-day gathering in Boston on November 4 where founders across all stages connect with top VCs and experienced entrepreneurs to gain tactical insights on building and scaling their companies.
Hello everyone. This episode features the most important and latest news about artificial intelligence and technology. GLT OPUS 5.5 has reached a new level. The model outperforms Fable 5.1 and is already competitive with GPT6 Astra in some tests. OpenI and Antropic discussed an…
— радуемся и гордимся за наших слонов! Погнали 🥳 🏆 3-е место, золотая медаль и 8000$ призовых на международных соревнованиях по Kaggle Это все забрал Максим Хандусь, студент 1 курса Хаба, вместе с руководителем направления по разработке нейронных сетей в Альфа банке Ильей…
Российский стартап Kodacode обогнал Claude Code в бенчмарке харнессов для DeepSeek.
🌟 Nvidia открыла веса модели для разметки спикеров в аудио Nemotron 3 Diarization – открытая модель диаризации на 100 млн параметров, которая работает на аудиозаписях и в потоковом режиме, размечая на записи отрезки речи каждого участника, включая моменты, когда люди перебивают…
Австралияның премьер-министрі OpenAI-ге үкімет бұзғаны туралы хабарлады және тергеуді бастады.
📉 Акции китайских ИИ-компаний упали до 12% после начала расследования в отношении DeepSeek и Moonshot AI 23 сентября распродажа затронула почти весь сектор: бумаги Z.AI потеряли 12%, MiniMax — 6,8%, Alibaba — около 5%, Xiaomi — 3,8%, Tencent — около 3%. Причиной стали сообщения…
📌 ИИ-агентам начинают писать инструкции — почти как новым сотрудникам Ещё недавно от корпоративного ИИ требовалось в основном одно: хорошо отвечать на вопросы по документам и базам знаний. Теперь от него хотят другого — чтобы он мог сам довести задачу до результата. Например,…