News
Microsoft Research Releases Webwright: A Terminal-Native Web Agent Framework That Scores 60. 1% on Odysseys, Up from Base GPT-5. 4's 33. 5%
1+ hour, 10+ min ago (396+ words) Webwright separates the agent from the browser and treats the browser as something the agent can launch, inspect, and discard while developing a program. The persistent artifact is not the browser session but the code and logs in the local…...
NVIDIA AI Releases Gated Delta Net-2: A Linear Attention Layer That Decouples Erase and Write in the Delta Rule
2+ hour, 24+ min ago (860+ words) Linear attention replaces the unbounded KV cache of softmax attention with a fixed-size recurrent state. This cuts sequence mixing to linear time and decoding to constant memory. The hard part is not what to forget. It is how to edit…...
Tencent Open-Sources Tencent DB Agent Memory: A 4-Tier Local Memory Pipeline for AI Agents
14+ hour, 35+ min ago (908+ words) Tencent has released Tencent DB Agent Memory, an open-source memory system for AI agents. The project ships under the MIT license. It targets a problem familiar to anyone shipping long-horizon agents: context bloat and recall failure. It is symbolic short-term…...
Build a Super Claude Framework Workflow with Commands, Agents, Modes, and Session Memory
15+ hour ago (237+ words) We install the required Python packages and import the core libraries needed for the tutorial. We initialize the Rich console so that our Colab output looks clean, structured, and readable. We then clone the Super Claude Framework repository and make…...
Nous Research Releases Contrastive Neuron Attribution (CNA): Sparse MLP Circuit Steering Without SAE Training or Weight Modification
23+ hour, 34+ min ago (862+ words) CNA requires only forward passes " no gradients, no auxiliary training, no iterative search. You define two sets of prompts: You run all prompts through the model. At each MLP layer, the method records down projection activations at the last token…...
A Step-by-Step Coding Tutorial to Implement GBrain: The Self-Wiring Memory Layer Built by Y Combinator's Garry Tan for AI Agents
1+ day, 15+ hour ago (408+ words) This is a hands-on tutorial. You'll install GBrain locally, import a small notes folder, run a real search, watch the knowledge graph wire itself, and connect it to Claude Code via MCP. About 20 minutes start to finish. All terminal outputs…...
Microsoft Releases Fara1. 5: A Family of Browser Computer-Use Agents (4 B/9 B/27 B) That Outperform Open AI Operator and Gemini 2. 5 Computer Use on Online-Mind2 Web
2+ day, 1+ hour ago (684+ words) Microsoft Research's AI Frontiers lab released Fara1. 5. It is a family of computer-use agent (CUA) models for the browser. The release ships three sizes: Fara1. 5-4 B, Fara1. 5-9 B, and Fara1. 5-27 B. The models are integrated with Magentic Lite, Microsoft's sandboxed browser interface for these agents....
Build Recurrent-Depth Transformers with Open Mythos for MLA, GQA, Sparse Mo E, and Loop-Scaled Reasoning
2+ day, 2+ hour ago (916+ words) In this tutorial, we explore Open Mythos by building an advanced recurrent-depth transformer workflow that runs end-to-end in Google Colab. We create both MLA and GQA model variants, compare their parameter counts, and check the stability of the recurrent injection…...
Google News
2+ day, 10+ hour ago (398+ words) Give Copilot Kit a " on Git Hub The framing that once pitted MCP against A2 A against AG-UI has given way to a recognition that these protocols solve fundamentally different problems " analogous to how TCP, HTTP, and HTML operate at different…...
Qwen Introduces Qwen3. 7-Max: A Reasoning Agent Model With a 1 M-Token Context Window
2+ day, 11+ hour ago (815+ words) Most AI models today are not designed for sustained, multi-step autonomous execution. Tasks like running hundreds of iterative code modifications, or chaining tool calls across hours without human intervention, require a different kind of model architecture and training focus. Alibaba's…...