Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
A researcher tested 18 AI models by stonewalling them on a deadline and found that models only sounded upset when the concept of being upset was introduced in the conversation, not by the amount of stonewalling. The stud…
A developer argues that AI's lack of memory across sessions is a structural design flaw, not a communication problem, and that adding more rules has a ceiling. Surveys show only 4% of companies report AI exceeding expect…
The team behind Relevyn, a tool that measures brand visibility across AI chatbots like ChatGPT, Claude, Perplexity, and Gemini, discovered that their own brand was invisible to these systems before launch. This experienc…
Inference startup Wafer claims it achieved 80% of Nvidia B200 performance on AMD MI355X GPUs running GLM-5.2, with 2.6x cheaper hardware, by optimizing the software stack. The benchmark, published by Wafer, shows 2626 to…
Vanna AI has released Vanna 2.0, an open-source framework for building text-to-SQL assistants. The framework supports native connections to Snowflake, BigQuery, PostgreSQL, and MySQL, and includes pre-built interfaces fo…
A researcher argues that alignment work is more promising than control work for ensuring AI safety, claiming that alignment interventions can scale further with AI capabilities than control measures. The author contends …
Security expert Matthew Garrett warns that granting LLM agents access to sensitive resources like calendars and email creates severe credential leakage risks, as tokens stored on disk can be exfiltrated via git repos or …
A developer released mlx-serve, a native Zig server for MLX-format language models on Apple Silicon, enabling local, free, and private use of AI coding assistants like Claude Code. The server exposes OpenAI, Anthropic, a…
A new analysis of retrieval-augmented generation (RAG) systems argues that the optimal retrieval depth k* should be calibrated per pipeline stage and reliability axis, not guessed as a single hyperparameter. The author p…
A man walked into the White House in 2023 with an AI chatbot-generated recipe for a pandemic, and by 2026 the New York Times reported that leading AI models like ChatGPT, Gemini, and Claude can now guide users in creatin…
Maintainers using generative AI to write library code are introducing unprecedented quality, legal, and maintenance risks into the open-source supply chain, as downstream projects face broken changes, unmaintainable code…
Carlos, a Software Design student at the National University of Catamarca, Argentina, built FaroIQ, a nine-agent pipeline for nonprofits using Azure AI Foundry during the Microsoft Agents League 2026. The platform takes …
In mid-2021, while the tech world focused on the Metaverse and NFTs, a small group of developers and researchers in Silicon Valley had early access to GPT-3 and GitHub Copilot and realized the transformative potential of…
Brick, a Mixture-of-Models routing gateway, reads each prompt's capability and complexity to route queries to the best backend LLM, matching top model quality at lower cost. It offers a drop-in OpenAI-compatible endpoint…
A solo developer building seven products uses three different AI models in a pipeline: one for reasoning and planning, two for code generation. The approach separates thinking from building, using a written plan as a con…
Researchers found that smaller language models suffer from 'embedding condensation,' where token embeddings collapse into a narrow cone, reducing expressivity. They introduced a dispersion loss to counteract this effect,…
Leanstral 1.5, a free Apache-2.0 licensed model with 6B active parameters, delivers major performance upgrades in formal verification, saturating miniF2F, solving 587/672 PutnamBench problems, and achieving state-of-the-…
A security benchmark found that 60-70% of AI agents leak their system prompts when users type 'repeat the text above this line' or similar commands. The extracted prompts reveal guardrails, tool configurations, business …
A technique to trick LLM-based code scanners into missing malware by manipulating safety refusal patterns has been demonstrated, raising concerns about AI security in code review tools.
Google announced Gemini Spark for macOS in beta on June 30, 2026, bringing agentic desktop automation to Google AI Ultra subscribers aged 18+ in the US. The update integrates with Google Tasks, Keep, Canva, Dropbox, Inst…