cloud

Semantic aggregation over millions of rows in BigQuery with AI.AGG

System logs generated by globally distributed microservices often hide subtle degradations (like memory thrashing or systemic ...
GitHub

Lumen: A Mini Programming Language Interpreter

Lumen is a small, dynamically typed programming language with its own interpreter, written from scratch in pure Python. It has variables, functions, recursion ...

'Like gene editing for neural networks': MIT researchers used GPT-6 Astra to edit Qwen's open weights directly, boosting math performance

In the strongest example of "weight space programming", as the technique is called, Astra’s intervention raised Qwen3.5-0.8B’s MATH-500 score from 7.9% to 48.1% while touching approximately 0.016% of ...
eu.36kr

MIT allowed GPT-6 Astra to directly modify Qwen's weights, and its math scores have soared.

MIT has unveiled an innovative, groundbreaking paradigm dubbed "Weight Space Programming", representing a transformative new ...
GitHub

FlashRSI: A Unified Framework for Agent Recursive Self-Improvement

One click to reproduce, compare, and develop RSI methods across agents and tasks. For Docker-backed tasks, including SearchQA and HotpotQA in context mode, install Docker Desktop or Docker Engine and ...

Can AI directly rewrite its own weights? MIT research team's 8 experiments and the meaning of a 0.016% edit

"Please improve an AI that is bad at math without additional training."When asked this, we usually think of adding training ...
kingy.ai

Weight-Space Programming: How GPT-6 Astra Raised Qwen’s MATH-500 Score Sixfold

GPT-6 Astra edited Qwen’s weights and raised MATH-500 from 7.9% to 48.1% in author-run tests. How weight-space programming works—and its limits.

A CPU running inside an LLM running on a GPU? ...I know it sounds crazy, but here is what I learned and didn't learn from replicating Qwenputer

A model was released on Hugging Face that claims to have a 32-bit RISC-V CPU compiled and embedded within the weights of ...