Skip to content
AI NEWS JOURNAL
AI, IN PERSPECTIVE

Search AI news

Search headlines, summaries and articles to connect the dots.

Newest first

v1.0.0

The llama.cpp project has released version 1.0.0, marking a significant milestone with numerous improvements across multiple platforms including CUDA, Vulkan, Metal, SYCL, OpenCL, and WebGPU. Key updates include enhanced support for quantized models (Q5_0, Q5_1, FP8), optimizations for flash attention and MoE caching, improved server security features like CORS protection and SSRF mitigation, UI enhancements such as sidebar reorganization and live-preview markdown editing, and better cross-platform compatibility. The release also includes bug fixes, CI/CD improvements, and updated dependencies.

What Do Tokenizer Comparisons Measure? Disentangling Granularity, Root Assignment, and Word-Level Readout across Amharic and Hausa

arXiv:2611.14818v1 Announce Type: new Abstract: Morphology-aware tokenizers are expected to help morphologically rich languages, and their reported gains over byte-pair encoding (BPE) are usually credited to their linguistic units. A comparison with BPE cannot support this attribution, since such a tokenizer also segments text more finely (granularity) and, on token-level tasks, changes how much the sub-token readout matters. We separate both factors from the effect of grouping words by their roots (root assignment) in Amharic, a Semitic language whose Ge'ez abugida hides consonantal roots inside syllabic glyphs, with Latin-script Hausa as an orthographic contrast. Our design combines a lossless root-and-pattern tokenizer with controls that change one factor at a time, in two independent implementations. In both, real root assignment improves Amharic named-entity recognition over scrambled roots (+4.9 and +4.7 F1), and root-frequency structure explains little of this gain. In small from-scratch models, the readout choice moves F1 by up to 13.4 points and reverses the sign of every comparison of the root-and-pattern arms with BPE, whereas in two pretrained encoders it moves F1 by less than one point. The granularity effect changes sign across languages and implementations. We recommend decomposing tokenizer effects, and reporting the readout, before a gain is attributed to linguistic structure.

When Is Proof-Step Verification Worth Its Cost? Operating Requirements of Certify-or-Decline Verification for Open-Weight Mathematical Reasoning

arXiv:2611.14794v1 Announce Type: new Abstract: Proof-step verification rewrites a solver's answer into typed steps and uses LLM judges to certify or decline. We evaluate this pipeline on open-weight models across GPQA, AIME-2025, and a MATH-500 subset. Certification raises answer-key-match precision by 31 percentage points on the strong GPQA configuration and 46 on AIME, while answering only 35% and 21% of attempts. On near-ceiling MATH-500, gains are at most 5 points at a loss of 13-27 points of coverage. An output-token-matched self-consistency selector obtains 82.9% precision at 35% coverage, compared with the verifier's pooled 87.6% (paired difference: +4.8 points, 95% interval [-8.2, 18.5]). A diagnostic sample reveals a distinct failure mode: 52 of 88 completed problems end in invalid proof construction without ever reaching a judge. Finally, external-LLM acceptance of rendered certified chains is 61.0-67.6%, compared with 87.6% answer-key match. These audits lack human or formal-kernel ground truth. Open-weight auditor pairs show near-zero chance-corrected agreement, which measures consistency rather than accuracy against proof truth. These results motivate separate reporting of execution failures, selected-answer precision, and independently assessed proof validity. We provide per-configuration results and evaluation details in the supplement.

Accelerating generative AI delivery with reusable assets

Every generative AI team keeps rebuilding the same components: RAG pipelines, document processors, and agent scaffolding. This post shows how to access a curated catalog of 23 production-ready, open source reusable assets from the AWS Generative AI Innovation Center, with real delivery results from DoorDash, Lyft, and Ricoh.

Automating AWS Agent Registry updates with push-based architecture

This post shows how to keep the AWS Agent Registry current automatically. Using AWS CloudTrail, Amazon EventBridge, AWS Lambda, and AgentCore Identity, an event-driven pipeline detects MCP server changes and pushes updated tool metadata to the registry with no stored secrets, across single- and multi-account environments.

Build autonomous AI agents with Amazon Bedrock and Spring AI

Learn how Java developers can build production-ready autonomous AI agents using Amazon Bedrock and Spring AI. This post walks through tool calling, conversation memory, resilience patterns, and multi-agent orchestration, then deploys the agent to AWS with Amazon Bedrock AgentCore.

Clipto enables ChatGPT and Claude to search and analyze years of people’s meetings and video calls

Artificial intelligence assistants such as ChatGPT and Claude have become essential tools for millions of workers worldwide, helping them compose emails in seconds, generate complex code and draft reports by looking at whatever data that person has access to. But there’s one thing that’s missing from those chatbots’ knowledge. Ask them what was discussed in […] The post Clipto enables ChatGPT and Claude to search and analyze years of people’s meetings and video calls appeared first on SiliconANGLE.

From prototype to production: Engineering agents with Agent DLC

Agents stall between demo and release when teams lack a repeatable way to measure how good an agent is and whether the last change helped. This post covers three agent properties that break traditional CI/CD, a measured case where an aggregate metric hides a coverage gap, and Agent DLC: one lifecycle, two loops, and one explicit standard for judging every change.

Hatz AI raises $24M to help small businesses go AI-native through their MSPs

Large enterprises have emerged as the foremost adopters of artificial intelligence technology, and it’s giving them an unfair advantage over their smaller competitors. Looking to change that is a startup called Hatz AI Inc., which has just raised $24 million in early-stage funding to empower managed services providers with the tools they need to keep […] The post Hatz AI raises $24M to help small businesses go AI-native through their MSPs appeared first on SiliconANGLE.

Inkling 2 Small now available on AI Gateway

Inkling 2 Small, a multimodal reasoning model supporting coding agents, visual understanding, and professional knowledge work, is now available on AI Gateway. It accepts text and image inputs and produces text output, enabling extended coding tasks, tool use for information gathering, document generation, and visual analysis including reading text in images, interpreting charts, and inspecting interfaces. The model can be used via the Vercel CLI with the identifier 'thinkingmachines/inkling-2-small' and integrated into tools like Claude Code, Codex, and Cursor. AI Gateway offers a unified API with features such as usage tracking, cost monitoring, retry configurations, failover, performance optimization, custom reporting, Zero Data Retention, budget controls, and routing rules. Pricing reflects provider rates with no markup or platform fees, including for Bring Your Own Key requests.

Inkling 2 Small on vLLM: Day-0 Support

vLLM supports Thinking Machines' Inkling 2 Small from day 0: attention residuals, a new decode kernel for relative-bias attention, all-reduce fused with the ops around it, and DSpark speculative decoding.

n8n@1.123.95

Release notes for n8n version 1.123.95, including a guided diff and auto-fix options via Cubic.

n8n@2.47.5

The n8n 2.47.5 update addresses a bug that allowed edits to read-only credential nodes via the Public API and fixes alignment issues with animated node borders in the editor.