AdvertiseWatchSubscribe
Updated , rescanned every two hoursMonday, October 26, 2026

▲ 218 today1.1k starsPythonHN ×1New · 5d old

Neuron Explorer lets you poke Qwen3.5's hidden switches

LLMsInference

What is neuron-explorer?

It loads the Qwen3.5-4B model, pushes each MLP neuron up and down on a few prompts, and saves every rollout. An interactive 3-D UI shows the neurons as points you can hover, click, and even steer in a live chat.

Who it is for

Prompt engineers and model-interpretability researchers can explore and steer individual neurons to diagnose behavior or design targeted prompts.

Get the signal first.

Join the list for the morning email: the day's biggest AI launches, explained in five minutes. Unsubscribe any time.

Prefer a reader? Follow the RSS feed.