# Apr 23, 2025

## ClickHouse gets lazier and faster: Introducing lazy materialization
- **ClickHouse's new lazy materialization** optimizes query performance by deferring column reads until necessary, achieving a **1,576× speedup** in specific scenarios without altering SQL code.

## π0.5: A VLA with open-world generalization
- The **π0.5 model** represents a significant advancement in **open-world generalization** for robots, enabling them to perform tasks in entirely new environments, such as cleaning kitchens and bedrooms not seen during training.

## Teaching LLMs how to solid model
- **LLMs can generate CAD models for simple 3D mechanical parts**, with ongoing research into training generative models on vast datasets of existing CAD files, enhancing their capabilities in design and manufacturing processes.

## [R] One Embedding to Rule Them All
- **OmniSearchSage** revolutionizes traditional two-tower architectures by employing a unified query embedding that enhances retrieval of pins, products, and queries through **multi-task learning**.

## Subnanosecond Flash Memory
- **Subnanosecond flash memory** has been achieved using a **two-dimensional (2D) Dirac graphene-channel** mechanism, demonstrating a program speed of **400 picoseconds** and robust endurance over **5.5 million cycles**.

## Can a single AI model advance any field of science?
- **AI's transformative potential** is being harnessed at Los Alamos National Laboratory (LANL) to tackle complex scientific challenges, with projects like ArtIMis aiming to develop foundation models that can generalize across various scientific domains.

## The Future of MCPs
- **MCPs (Model Context Protocols) are poised to revolutionize AI by acting as standardized APIs**, enabling large language models (LLMs) to access real-time data and perform actions, thus transforming them from static chatbots into dynamic agents.

## [R] [DeepMind] Welcome to the Era of Experience
- **AI is entering a transformative phase** where agents will achieve superhuman capabilities by learning predominantly from _experience_, rather than relying solely on human-generated data.

## MCP on AWS Lambda with MCPEngine
- **Model Context Protocol (MCP)** is emerging as the standard for enabling LLMs to interact with external tools, yet existing implementations often lack production readiness, particularly in stateless environments like **AWS Lambda**.

## PyTorch 2.7 Release
- **PyTorch 2.7** introduces **support for NVIDIA Blackwell GPUs** and **torch.compile** for custom operations, enhancing flexibility and performance in model execution.

## Making Brain Waves: AI Startup Speeds Disease Research With Lab in the Loop
- **BrainStorm Therapeutics** leverages **AI-powered computational drug discovery** and **organoids** to accelerate the development of treatments for neurological disorders, aiming to reduce the **93% clinical trial failure rate** associated with traditional methods.

## LongMamba: Enhancing Mamba's Long Context Capabilities via Training-Free Receptive Field Enlargement
- **LongMamba** enhances Mamba's long-context capabilities without training by addressing the limitations of **global channels** in state space models (SSMs), which struggle with memory decay as input lengths increase.

## Cost-Effective Text Clustering with Large Language Models
- **TECL** is a novel framework that leverages feedback from **large language models (LLMs)** to achieve accurate text clustering while minimizing the number of queries, thus addressing the **financial and computational overhead** associated with traditional methods.