All published intelligence, sovereign stack research, ecosystem news, and AI foundation model reviews.
Китайская модель GLM-5.3 уже почти на уровне Mythos. И это говорю не я — это говорит сама Anthropic. По крайней мере, если речь идёт о кибербезопасности. В статье по безопасности Anthropic пишет, что GLM-5.3 представляет угрозу из-за своих высоких возможностей и того, что модель…
🚨 Сотрудники OpenAI предупреждали руководство о проблемах с безопасностью ИИ ещё до серии инцидентов — но компания якобы не стала усиливать контроль По данным NYT, за несколько месяцев до взлома Hugging Face и других сайтов двое сотрудников сообщили топ-менеджерам, что новые…
NVIDIA Nemotron 3.5 ASR адаптируется для диалектов арабского языка в Саудовской Аравии.
Perplexity Research и turbopuffer анонсировали модель pplx-embed-v2-context-9b-preview для RAG-пайплайнов.
🌍 Freedom расширяет бизнес в Европе: во Франции — новые возможности, на очереди Швейцария Freedom24 расширяет присутствие в Европе. Французское подразделение компании получило статус связанного агента Freedom Finance Europe — теперь команда в Париже сможет…
Large language models (LLMs) are frequently updated for various use cases, where filtering out misaligned training samples is a common practice for preventing post-update misalignment. However, alignment is inherently context-dependent: a recommendation that is aligned in one…
A harness is the code around a language-model agent that organizes prompts, calls tools, manages context, and controls execution. As models grow stronger, recent work has begun to let agents improve their own harnesses, a line of work known as self-evolving harnesses. In most…
Voice agents in production must handle several requests, background speech, and customers who lose patience. We introduce VAmoS Energy, a benchmark that combines these challenges in 100 calls about utility billing and payment assistance. Each caller makes two to four requests.…
Language agents must revise planned actions when evidence changes, permission is revoked, or a stop instruction arrives. A useful response is selective: suspend affected actions, preserve unaffected work, and resume only after sufficient repair. We introduce NAQD-Env, a…
We study the generation of agentic workflows that jointly optimize multiple objectives, such as accuracy, cost, latency, robustness, and consistency. Existing methods for workflow generation typically optimize accuracy alone or a weighted sum of objectives, so each trained…
When a materials LLM answers a question about crystal structure, does it reason from the structure or copy an answer already printed in its input? Accuracy cannot tell: a structural description often prints the very field it is scored against. CARAT holds question and gold…
Personas used to seed LLM social simulations face a cold-start problem: existing methods lack a principled basis for deciding which attributes to include and how to assign their values. As a result, synthetic populations may misrepresent the demographic composition, latent…
Large language models (LLMs) can influence people's beliefs, yet little is known about whether and how they can manipulate each other. To investigate this, we simulate conversations between two agents: a target LLM that role-plays a human persona based on demographic and…
Vision-language models may rewrite anomalous text in images into linguistically plausible expressions, compromising OCR transcription faithfulness. Sequence-level task rewards and local teacher guidance are complementary, but guidance from the same teacher may not remain equally…
We present AREX-2, an effort to advance the self-improving capability of LLM agents, which we define as the ability to iteratively refine a solution at test time. This ability rests on two complementary capabilities: reflection, which produces a solution better than the current…
When a student is stuck, a tutor faces the assistance dilemma: help given too early can hinder productive struggle, while help withheld too long leaves the student in a frustrating, persistent impasse (i.e., wheel spinning). Generative AI tutors increasingly use guardrails…
High quality synthetic data is central to post training LLMs for adaptive AI applications that represent the diverse expert strategies and decisions in conversations. Prompting LLMs directly or conditioning them on end use scenarios yields low diversity data that collapses onto…
We study generalized Blaschke curves as a controlled environment for AI-assisted mathematical rediscovery. For one fixed degree-four Blaschke product, an agent receives numerical coordinates of the six pair-lines determined by each of 80 boundary configurations. The target…
Supervised fine-tuning of discrete diffusion language models masks some response tokens and trains the model to recover their original values from the visible context. The masking pattern therefore determines both the context available to the model and the tokens it learns to…
Concurrent autoregressive inference creates a fundamental interference problem: prefilling a newly arrived long prompt can delay tokens for requests that are already decoding. Fixed prefill chunks reduce this interference, but the best chunk size depends on the model, hardware,…
Leading AI developers have reported agents acting beyond their approved limits, which a United Nations panel described as an early warning of loss of human control. Yet incident reports and agent-safety evaluations describe these events differently, making it difficult to…
Virtual clients offer a cost-effective approach to support applications such as A/B testing, recommender system development, and interface evaluation. However, building them requires access to large-scale, semantically faithful, fine-grained online user trajectories. These data…
Recent progress in large language model reasoning has been driven by benchmarks and reinforcement learning environments with automatically verifiable rewards, particularly in mathematics, code, and formal logic. These settings make model accuracy easier to evaluate and optimize,…
Clinical large language model (LLM) agents deployed locally can consult more capable remote models, but doing so risks exposing patient information. Privacy-conscious delegation places disclosure decisions with a local agent, yet removing explicit identifiers is insufficient:…
Frontier LLMs are increasingly used to automate scientific research through iterative search. We distinguish idea-driven search from solution-driven search and identify three core challenges: organizing evolving research ideas, selecting promising directions, and maintaining…
Open-ended LLM homogeneity can create false plurality when several systems appear to offer independent perspectives while returning the same familiar default. Single-pass answers obscure the distinction between agreement produced by a tightly constrained answer space,…
This paper examines human-AI interactions (HAIIs) in clinical trials and presents a multidimensional categorisation framework that classifies interactions according to AI tasks, human-AI relationships, interaction configurations and interacting human groups. We define HAII,…
Large language models are frequently proposed as a route to AI-powered services for African communities, but they are least reliable exactly where the need is greatest: all African languages remain low-resource by any standard measure, and models trained on scraped, standardised…
While modern language models demonstrate impressive generative capabilities, they often struggle with constrained, counter-intuitive creative tasks. To address this limitation, we explore chess puzzle generation as a rigorous testbed for computational creativity and reasoning, a…
Large language models (LLMs) are increasingly used in culturally sensitive settings, where alignment requires representing diverse preferences within populations. Yet existing methods model populations at coarse demographic or community levels and overlook within-group…