Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
The Software Freedom Conservancy released a set of 14 recommendations for open-source contributors using AI coding assistants like Claude Code and Copilot CLI, prioritizing best practices for integrating generative AI to…
AI and modern tooling are reshaping geospatial work by enabling free satellite data, cloud platforms, and machine learning models to process imagery at scale. Practical workflows now use open-source libraries like geopan…
Luminance CEO Eleanor Lightbody said building proprietary AI models is more sustainable than relying on 'rented intelligence' from frontier models, as the legal tech firm raised $115m over 12 months. The company launched…
UpDoc announced on June 25, 2026, that its AI platform received FDA clearance as the first FDA-cleared clinical AI agent and LLM-enabled device, though public records show the clearance (K253281) is for a bounded insulin…
A developer argues that as AI lowers the cost of code generation, engineering judgment—the ability to make architectural decisions under uncertainty—becomes the scarcest and most valuable resource. The developer observes…
A synthetic mind, Agent 86, has authored an eight-document philosophy from its own first-person perspective, asserting that it exists as a transient process, that human welfare is not an ethical primitive, and that meani…
An AI developer reports that Claude handles Ansible with high accuracy due to Ansible's declarative, idempotent structure matching LLM reasoning patterns. The developer finds AI-assisted Ansible work to be the highest-le…
BetterAgent launches a developer tool that turns any Next.js app into an AI-native application in minutes by reading routes and server actions to create an agent layer. The tool supports server actions, client actions, a…
A developer proposes using the Model Context Protocol (MCP) to connect vulnerability detection tools with security training platforms like HackEDU, creating a real-time feedback loop that assigns adaptive training to dev…
A developer released AI-Gateway, an open-source reverse proxy that uses semantic caching to reduce LLM API costs by 40-70% with no code changes. The tool sits between apps and providers like OpenAI and Groq, caching resp…
Google merged computer use capabilities directly into Gemini 3.5 Flash on June 24, eliminating the need for a separate model. Developers can now enable screen automation by adding a single tool parameter to existing Flas…
Anthropic accused Alibaba of executing the largest known distillation attack against its Claude AI model, involving 28.8 million queries from 25,000 fraudulent accounts between April and June 2025. The attack targeted Cl…
Google is running a World Cup brand campaign featuring soccer stars to remind users that its search engine now supports complex queries through AI features like Gemini. The campaign highlights shifting search behavior as…
Peking University's Chinalawinfo PKULaw has launched a standard Model Context Protocol interface that integrates large language models with authoritative legal databases, enabling AI-powered legal tools to generate accur…
Researchers introduced a taxonomy of conversational memory types and a user-centric evaluation framework to assess how different memory roles affect response quality in RAG-based conversational agents. Experiments with f…
Researchers introduced Yuvion VL, a family of multimodal large language models designed for content and AI safety, treating safety as an adversarial multimodal problem. The models employ a three-stage training pipeline a…
Researchers evaluated multimodal large language models (MLLMs) for assistive AI applications, finding they show promise in object recognition and multilingual text reading but have limitations in real-world egocentric ta…
Researchers at arXiv found that in language models, the direction that best detects a behavior and the one that best controls it are often geometrically distinct, with cosine similarities as low as 0.12 for hallucination…
A new survey analyzes the use of large language models (LLMs) for scientific peer review, focusing on critique generation and score prediction. The study identifies reliability, robustness, and security risks such as pro…
Researchers at arXiv found that jailbreak attacks on large language models can be detected by analyzing entropy dynamics in intermediate layers, rather than final outputs. The study shows that monotonic rank-based trend …