Linum AI releases P-JiT, a 2B pixel-space text-to-image model trained for $21,038, which removes the VAE to achieve 4x attention context reduction.
The preview, available in iOS 27.3 and other OS betas, allows apps to integrate with Siri AI's new capabilities like personal context and onscreen awareness before the spring App Store submission window.
Baseten launches Grid, Sandboxes, and Loops to extend its inference platform with routing, execution, and training capabilities.
The new atomic start() function admits a run and claims a hook in one step, using a domain-specific token as the idempotency key to ensure exactly one run owns the token.
The marketplace now hosts over 25 partner tools, with eligible enterprise customers able to apply committed Anthropic spend toward these new solutions.
Moonshot AI was previously the only provider for these models, but AI Gateway now routes requests to multiple Zero Data Retention partners.
The design validates redirect URIs at the authorization step rather than registration, allowing users to trust specific https addresses while blocking open-redirect attacks.
A three-week engagement used near-policy SFT and SDPO to lift the model's quality by 20.7 points over the base and 8 points over the frontier competitor.
NVIDIA has released an open-source framework that automatically generates ontologies from DB metadata to validate AI agent SQL.
Valued at retail prices, the cost of AI tokens consumed by Anthropic employees surpasses their base salaries, excluding stock awards.
The arXiv paper (2605.29713) offers a derivation-oriented guide connecting PCA, VAEs, diffusion models, and GANs for researchers and students.
Apache-2.0 licensed platform that allows coding agents to tune risk models via MCP and supports decision models that operate with few labels.
The company outlines three principles for using human-in-the-loop workflows to define quality, evaluate context, and create golden datasets for continuous agent improvement.
The private preview tool integrates with Claude Code, Codex, and Cursor, claiming to raise root-cause accuracy on ORCA-bench from 6.3% to 35.2% through learned, versioned context.
Off-balance-sheet AI-related obligations for major US tech companies nearly doubled in six months to $3.2 trillion, reaching 80% of their assets.
The model is a RoBERTa fine-tune trained on 1.2B synthetic samples and identifies PII based on surrounding context.
LangChain manages the data plane infrastructure while keeping traces, evals, and agent deployments in the customer's own cloud account.
Implementations like Stately Agent, LangGraph, and OpenAI's Symphony enforce explicit execution states and transition rules to manage agent workflows and supervision.
The proof-of-concept modeled a $1.945 billion Term Loan B, using AI to convert legal agreements into executable gculpy contracts for near-instant settlement.
Apple has expanded the supported languages for its redesigned Health app to 10 languages.
Self-hosting Kimi K3 costs roughly $0.25 per million tokens, significantly below the $1.13 API price, indicating healthy margins rather than unsustainable subsidies.
Researchers at UC Berkeley and UCSF developed a streaming system that produces the first sound of a sentence within about one second of detecting a brain signal, compared to an eight-second delay for a full sentence in their earlier approach.
AI2 introduces StateBench, a four-level benchmark revealing that while recurrent architectures generalize better than transformers, top models like Gemini 3.8 Flash and Claude Opus 5.5 still fail on hard state-tracking tasks.
Walker argues that the 2013-2016 Cyber Grand Challenge relied on expert systems rather than machine learning, and that current AI capabilities have disrupted the traditional balance of offensive scarcity.
AI Briefing