AI Squawk the real-time terminal for AI news

The issue affecting claude.ai and Claude Cowork has been resolved. — Anthropic Status +1 source

The issue affecting download performance on the AWS CDN in the Singapore region has been resolved. — Hugging Face Status

All affected services are operating normally again. — OpenAI Status

We've resolved the incident and the affected service is now operational. — Fireworks AI Status

We've resolved the incident and the affected service is now operational. — Fireworks AI Status

594B Mixture-of-Experts model (17B active) with image input and up to 1M tokens of context, built for agentic coding and knowledge work. — Ollama Library Inkling

We've resolved the incident and the affected service is now operational. — Fireworks AI Status

We've resolved the incident and the affected service is now operational. — Fireworks AI Status

At the start of November, Chris Phillips, Google's vice president and general manager for education, said K-12 schools will soon have to… — HN "AI" (>=75pt)

We've resolved the incident and the affected service is now operational. — Fireworks AI Status

Agent Anomaly Detection (AAD) provides asynchronous, low-latency monitoring for AI agents by analyzing OpenTelemetry traces and logs to… — Google Developers Blog

How pairing Go's surviving compiler bounds-check traps with Gemini 4 Argon and the Z3 constraint solver catches latent out-of-bounds panics… — Google Developers Blog

Language model (LM) alignment improves model outputs to reflect human preferences while preserving the capabilities of the base model. — Apple Machine Learning Research

fix: Preserve bounded JSON usage metadata and reuse OpenAI wire types; breaking types require narrowing usage details, Realtime image content, nullable MCP… — OpenAI Agents SDK releases (JS)

We've resolved the incident and the affected service is now operational. — Fireworks AI Status

Meet Edge0-145B, an open-source 145B model for local coding, reasoning, and agent tasks, with no cloud-model token fees for local inference. — HN "AI" (>=75pt)

Sum GPTBigCode's shared key/value gradient across tensor-parallel ranks by @Achyuthan-S in #8928 … — DeepSpeed releases

Agent skills capture repeatable engineering work in files teams can review and reuse, while tools, project context, and permissions remain… — GitHub Blog (AI & ML)

Our forensic analysis of publicly released AI models finds evidence that DeepSeek, Moonshot AI, MiniMax and Ant Group have gained access to… — HN "AI" (>=75pt)

tracing: turn off FastAPI's native tracing on Agent Engine (87e5399) — Google ADK releases (Python)

The surge of AI and AI-generated apps is overwhelming Apple’s app review process, leaving developers caught in the crossfire – two examples! — HN "AI" (>=75pt)

We've resolved the incident and the affected service is now operational. — Fireworks AI Status

We are currently investigating this issue. — Baseten Status

Add --flavor=interactions-eap packaging and scrub private exports (39f6abe) … — Google GenAI Python SDK releases

A detailed root cause analysis will be shared as soon as it is available. — GitHub Status

How multimodal language models should encode visual input remains up for debate. — arXiv (AI/CL/LG, keyword-filtered)

Model architecture shapes the scaling behavior, training efficiency, and inference cost of LLMs, making it a central focus of both industry… — arXiv (AI/CL/LG, keyword-filtered)

Applying machine learning systems to new domains introduces bottlenecks in both labeled data acquisition and manual pipeline engineering. — arXiv (AI/CL/LG, keyword-filtered)

Remasking allows diffusion language models to revisit generated text, but effective correction requires knowing when and what to revise. — arXiv (AI/CL/LG, keyword-filtered)

Context attribution asks how removing parts of a prompt changes a language model's output. — arXiv (AI/CL/LG, keyword-filtered)

Looped models have emerged as a prominent paradigm for reasoning, with successes on simple reasoning puzzles such as Sudoku-Extreme and… — arXiv (AI/CL/LG, keyword-filtered)

Local Binary Patterns (LBP) remain among the most widely used texture descriptors, from industrial inspection to medical imaging. — arXiv (AI/CL/LG, keyword-filtered)

We investigate whether and how LLMs encode goals: internal neural representations that precede and guide action, and that are flexible… — arXiv (AI/CL/LG, keyword-filtered)

On-device inference requires kernels that make effective use of local accelerators. — arXiv (AI/CL/LG, keyword-filtered)

Today, we're rolling out Gemini 4 Hydrogen inside Google Antigravity alongside a new model dropdown option called Orchestrator, which pairs… — HN "Gemini" (>=50pt)

We propose Synthesis Smuggling, a novel threat model in which an adversarial LLM infers sensitive information about a user from routine… — arXiv (AI/CL/LG, keyword-filtered)

We introduce GeoWorld, a controllable and continuously evolving geographic environment platform for LLM agents, and LiveGeo, a benchmark… — arXiv (AI/CL/LG, keyword-filtered)

A detailed root cause analysis will be shared as soon as it is available. — GitHub Status

All affected services are operating normally again. — OpenAI Status

Modern automated heuristic discovery (AHD) systems increasingly rival hand-crafted heuristics by leveraging Large Language Models (LLMs) to… — arXiv (AI/CL/LG, keyword-filtered)

In interactive systems where users engage with LLMs, users are adaptive and, after observing LLM responses, can change their prompting… — arXiv (AI/CL/LG, keyword-filtered)

This paper presents Al Ghasali, a computational framework designed to formalize and simulate Islamic jurisprudential (Fiqh) rulings. — arXiv (AI/CL/LG, keyword-filtered)

Oral exams are compelling because (1) students can be asked to provide their reasoning for an answer and (2) their interactive nature… — arXiv (AI/CL/LG, keyword-filtered)

Federated learning (FL) enables independent data owners to train models collaboratively without centralizing raw data. — arXiv (AI/CL/LG, keyword-filtered)

From 08:03 PT / 16:03 UTC on November 12 to 08:40 PT / 16:40 UTC on November 13, mail, calendar and search tools in the Microsoft 365 connector and Claude for… — Anthropic Status

Summary Agno v3.2 makes Super Agent faster and more capable with direct tools, expands Company task governance and coordination, introduces… — agno releases

Decision models answer a question with a typed output, such as a choice among options with a probability for each, so one model can choose… — arXiv (AI/CL/LG, keyword-filtered)

Retrieval-augmented generation systems attach citations to attribute claims to their sources. — arXiv (AI/CL/LG, keyword-filtered)