'than' in comparisons

What this neuron does

This neuron lights up on the word 'than' when it completes a comparison, like 'more awful than' or 'moodier than usual'. Turning it up made the model's text drift into unrelated topics, so the effect on the writing itself is hard to read.

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

   multilingualism.⏎While the most effective way to combat illegal online activities is to attack them at source, the multi-national environment of the Internet makes this more complicated than just shutting down a local server.⏎Filtering is not a solution - the real

2. #2 peak

   disorders did not seek treatment. About 15% of the respondents<|endoftext|>Okay, I am by no means knowledgable in this field, probably more dangerous than anything. But I have a question about gravity. If two masses are identical (

3. #3 peak

   separate genre or segregated strand of literature.⏎- aim to keep up to date with matters surrounding diversity and inclusion.⏎If you feel like we could be doing this better than we already are then please don't hesitate to send us your thoughts via the contact

4. #4 peak

   causes CAH.⏎An X-ray can help to diagnose CAH in children. Because some children with CAH grow too quickly, their bones will be more developed than<|endoftext|>grupa 1111 a - automatizari⏎artificial

5. #5 peak

   shortens. You may find yourself feeling moodier than usual and without the energy you normally have during the warm months to get things done

6. #6 peak

   went to the bathroom⏎- You might feel a burning sensation while you are urinating and you will also feel like you need to go to the bathroom more often than normal⏎- You will notice pain below your navel and will have a general

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

   appear and disappear.⏎Being responsible for a tiny person can be intense. A few tips for taking care of yourself when you have a lot going on:⏎Don't skimp on your own nutrition, and avoid too much

2. #2 peak

   by scalars and indicates why they are useful.⏎A short introduction to complex numbers written primarily for students aged 14 to 19.⏎This article only sk

3. #3 peak

   charge more due to changes in petrol prices and the crude price. Due to the state taxes every region price is set accordingly<|endoftext|>Links / Pages:⏎Page at a Gl

4. #4 peak

   - Other medications you may currently be taking.⏎As with any medication, do not adjust your dose unless your healthcare provider specifically instructs you to do so.⏎(Click Monopril Dosage for more information about dosing guidelines for this medication.)

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.