Liquid AI released two new retrieval models, LFM2.5-Embedding-350M and LFM2.5-ColBERT-350M, for fast multilingual search across 11 languages. The models, available on Hugging Face, are designed as dro…
Salesforce released a tutorial demonstrating an end-to-end workflow for its CodeGen model, covering loading from Hugging Face, generating Python functions from natural-language prompts, and adding val…
Perplexity launched Brain, a self-improving memory system for its Computer agent that builds a context graph of the agent's work and learns overnight to improve performance. Brain remembers what the a…
OpenAI released LifeSciBench, a benchmark of 750 expert-authored tasks evaluating AI models on real life-science research. The best model, GPT-Rosalind, scored 36.1%, indicating significant room for i…
NVIDIA released SkillSpector, a tool that scans AI skills for security risks using static analysis and SARIF reports. The tool evaluates skills through a LangGraph workflow, organizes findings with pa…
Vercel released eve, an open-source framework for building and scaling AI agents where each agent is defined as a directory of files mapped to capabilities. The framework, published as an npm package …
MiniMax released MSA (MiniMax Sparse Attention), a block-sparse attention method that reduces the quadratic cost of softmax attention at long contexts. The method was trained on a 109B-parameter Mixtu…
OpenAI published a new pre-deployment safety method called Deployment Simulation, which replays past conversations through a candidate model to estimate undesired behavior frequency before release. Th…
MarkTechPost published a tutorial on building memory-efficient Transformers using xFormers, covering packed sequences, grouped-query attention, ALiBi, SwiGLU, and causal attention. The guide demonstra…
The Qwen team released three embodied AI models—Qwen-RobotManip, Qwen-RobotWorld, and Qwen-RobotNav—as part of Qwen-Robot-Suite, each built on a Qwen vision-language backbone and targeting different r…
Nous Research has updated its open-source Hermes Agent to support asynchronous subagents, allowing delegated tasks to run in the background without blocking the parent chat. The update introduces a ne…
Atoms, a new vibe coding tool built by the team behind MetaGPT, uses a multi-agent AI system to handle the entire product lifecycle—from market research and app development to deployment and marketing…
Google Cloud launched the Open Knowledge Format (OKF), a vendor-neutral markdown specification designed to provide AI agents with curated context. The spec formalizes the LLM-wiki pattern using markdo…
Tokyo-based Sakana AI launched its first commercial product, Sakana Marlin, an enterprise research agent that autonomously generates up to 100-page reports with slides over eight hours. The agent uses…
Researchers from UC Berkeley and UT Austin released Flash-KMeans, an IO-aware, exact k-means library that runs over 200× faster than FAISS on GPUs by restructuring data movement. The open-source libra…
Z.ai released GLM-5.2, its latest large language model, featuring a usable 1-million-token context window and two thinking-effort levels (High and Max), but no benchmark scores at launch. The model, t…
Anthropic's Claude Code has evolved from a terminal coding assistant into a layered agentic system with distinct memory, hooks, skills, subagents, plugins, and MCP layers. A new guide details 25 featu…
A tutorial demonstrates streaming, filtering, deduplication, tokenization, and analytics on the FineWeb dataset using Python, reproducing quality-filtering pipelines and MinHash-based near-duplicate d…
Databricks released Omnigent, an open-source meta-harness for AI agents under the Apache 2.0 license. The tool provides a common interface above command-line agents like Claude Code, Codex, and Pi, en…
A new tutorial demonstrates how to build a QwenPaw agent workspace with custom skills, model providers, console access, and streaming API testing. The guide covers installation, authentication, worksp…