'to' or verb after 'what' or 'whether'

What this neuron does

This neuron fires on the word that comes right before a verb in a decision or intention phrase, such as the 'to' in 'what to believe', 'whether to enroll', or 'choose to take'. Turning it up made the model repeat phrases, and turning it down made it drift into maths word problems.

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

   deciding that Howard, the fittest of the three, should continue alone to find help, leaving Mick to care for Cleve who by then,

2. #2 peak

   . It means analyzing the choice of what to believe or do. You must be thinking, how is it different from the usual

3. #3 peak

   needed. While it may sound trivial, the decision on what to produce/grow has important ramifications. If you plan to grow/produce a new

4. #4 peak

   -- Umberto Eco⏎More information about the Python-list<|endoftext|>Jayne Davis: “I am not convinced in my area of research there are less females choosing to take their career in this direction.”⏎We spoke to Jay

5. #5 peak

   and this comes from close observation of the plants plus a large application of common sense. Just take one step at a time. Master the fundamentals before you succumb to the excitement of special techniques or the challenge of difficult subjects

6. #6 peak

   were to succeed each other, moving up the hierarchy of cities toward Kiev, filling vacancies left by the advancement or death of an elder brother. The youngest brother was to be succeeded as grand prince by his eldest nephew whose father had been a grand prince

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

   outer ear infection.⏎You should tilt your head to the side so the affected ear looks up when applying the drops. To widen the ear canal:⏎- Gently pull the earlobe back and down.⏎- Before using the ear drops, give the

2. #2 peak

   oving tool. Grasp the tick as close to the skin as possible. If you don’t have tweezers, protect your fingers with a tissue.⏎- Pull the<|endoftext|>Listen to radar echoes from satellites and meteors, live on listener

3. #3 peak

   were transported to the shore. The press was desperate to find out what exactly happened out there.⏎Thousands of people gathered on the streets of New York, patiently waiting for the arrival of RMS Carpathia…⏎This doesn’t come as a surprise,

4. #4 peak

   was standing. I managed to move behind the trunk of a spruce tree. My trembling fingers clutched the cannister of bear spray fastened to my beltóa last resort in case of

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.