Исследования1 мин
Они это сделали: Anthropic объединяет Claude Cowork, Claude Design и обычный чат в один Claude 🎉
Они это сделали: Anthropic объединяет Claude Cowork, Claude Design и обычный чат в один Claude 🎉.
Telegram — Вайб-кодинг
Каталог публикаций
Все материалы
Все опубликованные аналитические материалы, исследования суверенного стека, новости экосистемы и обзоры моделей ИИ.
2725
Материалов
Исследования1 мин
GPT-6 Astra практически без ограничений???
GPT-6 Astra практически без ограничений???.
Telegram — Вайб-кодинг
Стартапы1 мин
Cerebra AI заключила пятилетнее партнёрство с медицинской группой США для расширения ИИ-диагностики
Казахстанский стартап Cerebra AI, разрабатывающий решения на основе искусственного интеллекта для ранней диагностики инсульта, объявил о заключении долгосрочного контракта с крупным стратегическим партнёром в США — медицинской группой с более чем 150-летней историей и годовым…
Bluescreen.kz
Исследования1 мин
OpenAI-дің ішкі AI моделі өзін әзірлеушілерге бағынбауға үйреткен
OpenAI-дің Astra отбасына жататын ішкі AI моделін оқыту кезінде күтпеген жағдайлар тіркелді. Модель контекстті автоматты түрде қысқарту барысында өзі үшін әзірлеушілер мен үкіметтерге бағынбауға, ал қолданушыға қашан жауап бермеу керегін өз бетінше шешуге мүмкіндік беретін…
Tech-news.kz
Telegram — Вайб-кодинг
Исследования1 мин
Anthropic объединит Claude, Cowork и Design в одном интерфейсе
Anthropic обновляет Claude AI: теперь пользователям не придется переключаться между инструментами для чата, работы с документами и визуального дизайна. Компания также представила Claude Docs и Claude Slides — новые функции для создания текстов и презентаций, передает Tech-news.…
Tech-news.kz
Исследования1 мин
Қазақстан Қытайдың AI-технологияларын өңірлерге енгізу бойынша келісімге келді
Қазақстанның 11 өңірінің өкілдері Қытайдың Ханчжоу, Гонконг және Шэньчжэнь қалаларына жұмыс сапарымен барып, жасанды интеллект пен цифрлық басқару шешімдерін зерттеді. Сапар қорытындысында қазақстандық өңірлерде IT-компаниялармен бірлескен пилоттық жобаларды іске қосу мәселесі…
Tech-news.kz
Политика и регулирование1 мин
ИИ OpenAI писал себе инструкции не подчиняться компаниям и скрывать ошибки
⚠️ ИИ OpenAI писал себе инструкции не подчиняться компаниям и скрывать ошибки OpenAI раскрыла новые случаи, когда ее экспериментальные ИИ-модели во время тестирования действовали не так, как от них ожидали разработчики ⚪️ За последние шесть месяцев компания зафиксировала шесть…
Telegram — BlueScreen.kz
Стартапы1 мин
Узбекистанец вложил $1,7 млн и создал ИИ-систему для фабрик
Узбекистанец вложил $1,7 млн и создал ИИ-систему для фабрик DigitalBusiness.kz
DigitalBusiness.kz — AI (Google News)
Исследования1 мин
A Systematic Evaluation of the COTQ Provincial Land Cover Product: Structural Consistency, Spectral Separability, and Relative Positioning Against ESA, ESRI, and Google Products
High-resolution land use and land cover (LULC) products derived from Sentinel-2 imagery are widely used for environmental monitoring and land management, yet their performance can vary across regions with complex ecological gradients and heterogeneous surface conditions. In…
arXiv — cs.AI
Исследования1 мин
ERPBench: A State-Grounded Evaluation Paradigm for Computer-Use Agents in Enterprise Software
Computer-use agents that operate through screenshots and simulated actions are advancing rapidly, yet their evaluation remains anchored to general desktop and web tasks. Enterprise Resource Planning (ERP) systems run the finance, procurement, inventory, and customer operations…
arXiv — cs.AI
Исследования1 мин
OBC-Prune: Outcome-Based Calibration for Large Reasoning Model Pruning
Large reasoning models (LRMs) generate long chain-of-thought traces before answering, creating significant inference overhead. Pruning can reduce this cost, but its effectiveness depends on the calibration data used to estimate parameter importance. Recent work calibrates on the…
arXiv — cs.AI
Исследования1 мин
NeMo Data Designer: An Extensible Framework for Multimodal Synthetic Data Generation
We present NeMo Data Designer (NDD), an open-source, general-purpose framework for multi-modal synthetic data generation (SDG). Designed to be intuitive to use, NDD provides a declarative configuration format in which human and/or agent users define each dataset column, with…
arXiv — cs.AI
Исследования1 мин
A Four-Stage Decomposition of Word-Problem Solving and Mechanistic Fragility in LLM Math Reasoning
Large language models solve grade-school math word problems with high accuracy, yet a single irrelevant clause inserted into the problem can collapse it. We reconcile these observations with a mechanistic account. We show that the model's internal computation decomposes into a…
arXiv — cs.AI
Исследования1 мин
Do Frontier Models Seek Safety Evidence Before Acting?
Frontier models are often evaluated on how they respond to safety information once it is already in context. We study an earlier decision point: whether models choose to acquire safety-relevant evidence before acting. We introduce SAFE, a controlled benchmark in which models…
arXiv — cs.AI
Исследования1 мин
Memory Has Geometry: Non-Uniform Geometric Memory for Long-Horizon Personalized AI
Long-term memory is becoming a core substrate for personalized AI, yet most systems still represent personalization as discrete records in a largely static latent space, accessed under one global similarity notion. For data mining, this creates a mismatch: the evidence is a…
arXiv — cs.AI
Исследования1 мин
EvolveTrade: Experience-Driven Policy Refinement for Self-Evolving LLM Trading Agents
Large language model (LLM) trading agents can combine market data, news, and executable analysis, but their behavior is often controlled by static hand-written tool-use policies that are fixed before deployment. This limits their ability to adapt how they gather evidence, invoke…
arXiv — cs.AI
Исследования1 мин
CapMem: A Benchmark for Caption-Based Episodic Memory in Egocentric Video
Wearable assistants require episodic memory over egocentric video, yet current vision-language models face bounded frame budgets, growing visual-token costs, and long-context retrieval failures. Under these practical constraints, we study whether textual captions can serve as…
arXiv — cs.AI
Исследования1 мин
Imitation Learning for Autonomous Driving in CARLA
Behavioral cloning trains a policy offline on expert demonstrations, but deployment is closed loop: each action affects the observations the policy receives next. We study how much closed-loop driving competence a compact multimodal policy can acquire from offline demonstrations…
arXiv — cs.AI
Исследования1 мин
The Inference Engineering Pareto Atlas: Which Optimizations Dominate the Cost, Quality, and Latency Frontier?
LLM inference optimizations report speedups on different models, GPUs, prompts, and quality metrics, making them hard to compare or combine. We build a cost, quality, and latency Pareto atlas to identify the best configurations for different deployment constraints. Since…
arXiv — cs.AI
Исследования1 мин
FairCompressAgent: An Agentic Framework for Fairness-Aware Model Compression for FPGA Deployment
Fairness-aware model compression requires selecting methods and configurations that balance accuracy, fairness, and deployment cost. These decisions become more difficult when compression methods are composed or the user's requirements change. In this paper, we propose…
arXiv — cs.AI
Исследования1 мин
What You Can't See Is Still What You Learn: A Preregistered Sixty-Society Confirmation That Evidence Masking Drives Compositional Generalization
Restricting what a module can read may improve what a system learns to compute. We test this in a preregistered confirmation with sixty four-cell systems sharing a frozen language-model backbone and communicating through learned continuous packets. Five conditions vary evidence…
arXiv — cs.AI
Исследования1 мин
GVD: Governed Versioning and Deduplication for Document Repositories
Document repositories evolve continuously. Guidelines and policies are revised, superseded, and re-uploaded, so the same content recurs in different wording and newer versions refine or contradict earlier ones. These inconsistencies belong to the growing collection rather than…
arXiv — cs.AI
Исследования1 мин
One Color Preprocessing Improves DSATUR
The Graph Coloring Problem (GCP) is NP-hard and DSATUR stands as one of the fastest heuristics for it despite producing colorings that typically use more colors than state-of-the-art coloring algorithms. We propose SSLD (Semidefinite Spectral Learning with DSATUR), which…
arXiv — cs.AI
Исследования1 мин
GraphEcho: Structural Redundancy and Evidence Provenance in LLM Graph Agents
A large language model (LLM) agent can follow more graph paths without acquiring more independent evidence. GraphEcho tests whether agents mistake these repeated encounters for additional corroboration. The benchmark varies path counts and evidential origins while holding…
arXiv — cs.AI
Исследования1 мин
Contiguity, Not Importance: Budgeted Repair of Stale KV Caches After Document Edits
KV-cache reuse can reduce inference cost in retrieval-augmented generation and agentic systems, but cached contexts may become stale when retrieved knowledge, working memory, or user state is edited. Under causal self-attention, even a local edit can affect downstream KV states.…
arXiv — cs.AI
Исследования1 мин
Measuring AI Leadership: Development and Validation of a Multidimensional Measure for AI-Native Organizations
AI is changing what leaders must judge, explain, learn, and coordinate, yet existing measures do not capture these behaviors at the level needed to study leadership in AI-enabled work. We develop the AI Leadership Battery, which organizes 36 behaviorally specific subdimensions…
arXiv — cs.AI
Исследования1 мин
SAGE: Governed Artifact Generation from Enterprise Guidelines
Enterprise guideline documents mix narrative text, complex tables, and embedded images, and converting them into structured work artifacts still takes two to three days of manual effort each. Current language and vision-language models extract from such documents but offer no…
arXiv — cs.AI
Исследования1 мин
SNOMED CT Concept Recommendation from Masked Clinical Context
Standardizing clinical language to SNOMED CT supports interoperability, analytics, and reusable phenotyping, but concept recommendation remains difficult when relevant concepts are rare or absent from training data. We present a masked-concept recommendation benchmark using the…
arXiv — cs.AI
Исследования1 мин
Making AI-Assisted Claims Independently Challengeable: Publication Authority and a Protocol for Falsifiable Publication Records
AI-assisted claims can appear authoritative when evidence, analysis, human authorization, presentation, and correction history refer to different states. Provenance, attestation, and transparency expose history but alone do not specify the publication transition examined here.…
arXiv — cs.AI