MIT Researchers Use GPT-6 Astra to Improve Qwen3.5 by Editing Its Weights

1 reports 1 reporting sources Updated 2 hours ago

Understand the event

AI overview

MIT researchers used GPT-6 Astra to inspect and edit a Qwen3.5-0.8B model, changing 0.016% of its weights without training or gradient descent. The model’s MATH-500 score rose from 7.9% to 48.1%, and its instruction-following performance also improved. In other experiments, Astra identified neural network functions, modified a car-game policy, and generated weights for a MinAtar network.

Generated by AI from reports · Updated 1 hours ago

Latest development11/3 09:57 (UTC+8)

MIT Researchers Show GPT-6 Astra Editing Qwen3.5 Weights Without Training QbitAI

Models in this story

GPT-6 Astra#3 overall$10 input / $50 output per million tokens

API prices of all models

Report timeline

Follow the reports to explore different sides of the event.

11/3

09:57
QbitAIPicks
MIT Researchers Show GPT-6 Astra Editing Qwen3.5 Weights Without Training
MIT researchers used GPT-6 Astra to inspect and edit a Qwen3.5-0.8B model, changing 0.016% of its weights without training or gradient descent. Its MATH-500 score rose from 7.9% to 48.1%, while instruction-following performance also improved. The article describes additional experiments in which Astra identified network functions, modified a car game policy, and generated weights for a MinAtar network.

Model releases, price changes and compute developments, with original sources. Free, confirmed by email, and easy to unsubscribe. Read a past issue

Event heat trend

There is not enough continuous observation data to plot a trend yet.