"to" after made or need-type verbs

What this neuron does

This neuron lights up on the word "to" when it introduces an action someone or something is forced, made, or needed to do, as in "made to work" or "need people to fight." It goes quiet, and slightly negative, in "beaten to death." Turning it up made the model's writing drift into unrelated rambling rather than anything about this pattern.

Written by Claude from this neuron's activations and steering outputs. Confidence: medium. Think it is wrong? Flag it (opens a prefilled issue on GitHub).

Fires most on

The text windows where this neuron's activation was highest. Amber marks positive activation; the outlined token is the peak. Hover a token for its strength.

1. #1 peak

   a Chemistry Tutor<|endoftext|>Those who fight for us, also need people to fight for them.⏎The military physically protects us and sacrifices for us, yet some

2. #2 peak

   Of International Conference On Communication Science⏎This work is licensed under a Creative Commons Attribution 4.0 International License.<|endoftext|>How do you explain how babies are made to an 8-year-old?⏎This topic has been the subject of countless jokes over

3. #3 peak

   at 18:48 BST⏎It is a form of modern day slavery otherwise known as forced labour. Millions of men, women and children are made to work against their will for little or no pay.⏎The UN's International Labour Organisation

4. #4 peak

   , liberty or possessions; for men being all the workmanship of one omnipotent and infinitely wise Maker… they are His property, whose workmanship they are made to last during His, not one another's pleasure.⏎This idea found its way into

5. #5 peak

   what counts as ‘health data’.⏎We put forward evidence that the datafication of health has created an ‘Internet of Health’, in which everyone is made to perform as a patient, subject to continuous monitoring, diagnosis and prediction. We ask

6. #6 peak

   The substrate is made to be the cathode and the coating the sacrificial corroding anode. An

Steering

Push this neuron up or down while the model writes, and watch what changes. The outputs were generated ahead of time, so nothing runs in your browser.

Most negative on

Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.

1. #1 peak

   the woman's story, and when the crowd took notice of a hearse parked outside of the store, the rumor began to circulate that Rivera had been beaten to death. The woman who had raised the alarm was arrested for disorderly conduct,

2. #2 peak

   Pandits fled the very next day. In April 1990, a Kashmiri Pandit nurse Sarla Bhat was gang-raped and beaten to death, in June Girija Tickoo was also gang-raped and then sliced

3. #3 peak

   .⏎Once again, Leah made my dream come to my life. Her chameleonic beauty makes it so easy to transform in every possible character, from Sleeping Beauty to Wonderwoman, to my very own Girl with a Pearl Earring.⏎To

4. #4 peak

   century it was burned to the ground and was rebuilt shortly afterwards only to be abandoned a century later, the

Which neurons in earlier layers feed this one, and which later ones it feeds. Amber links excite, blue links suppress; longer bars are stronger.

Activation histogram

How often this neuron fires at each strength. Almost always it does close to nothing (the tall bars in the middle); the snippets above are the rare moments it fired hard, amber for positive and blue for negative.

Logit lens

The words this neuron nudges the model toward (amber) or away from (blue) when it fires. A hint, not a label: many neurons work through later layers instead.