word-start 't' fragment in split words

What this neuron does

This neuron lights up on the letter t when it begins the first piece of a longer word that has been split in two, such as the t in tenses or tics. When it was turned up, the model's text drifted into repeated, broken phrases, which doesn't show a clear theme.

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

   the active verb) + Past Participle Example: A letter was written.⏎VERB TO BE - Verb to be in the passiveVerb to be in different tenses am-is-are +past participlePresent simple (am-is- are

2. #2 peak

   tense cohesiveness is important. However, understanding the different types of tense can make you, well, tense.⏎At the most fundamental level, there are three tenses: past tense, present tense and future tense. Each of these tenses

3. #3 peak

   transmit or reflect; "a white color is made up of many different wavelengths of light"⏎tinge, undertone - a pale or subdued color⏎4. tincture - (pharmacology) a medicine consisting of an extract in an

4. #4 peak

   especially about the personal or private affairs of others; the act is also known as dishing or tattling.⏎Gossip is a topic of research in evolutionary psychology, which has

5. #5 peak

   blinking the eyes or shrugging the shoulders. Simple tics can also involve vocalizations, such as grunting or clearing the throat repeatedly.⏎Additional examples of simple tics that may occur in Tourette syndrome include:⏎- Opening the mouth⏎-

6. #6 peak

   the side.⏎Some individuals who experience Tourette’s may engage in complex motor tics that involve some form of self-harm, such as hitting oneself. Complex tics can also involve actions such as touching items, hopping or tapping the feet.

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

   region of Kenya (January, 2009), a German canton of Switzerland (March, 2009), and the Piemonte-Toscana region of northern/central Italy (images only, February-March,

2. #2 peak

   first harvest and gathering as neighbors to give thanks.⏎As often happens when our children are home for the holidays, a vigorous discussion about truth, liberty, and sometimes-tumultuous American history ensued. But it got me thinking about the origin of

3. #3 peak

   while pursuing social policies geared towards economic recovery. David Cameron, Prime Minister of the UK, was elected in 2010 on the promise of a “tory miracle”. He promised a “country that leaves the European Union” will emerge

4. #4 peak

   ), strengthen the dorso-lumbar area and the abdominal strap.⏎STRENGTHEN YOUR JOINTS⏎It mobilizes and activates all the joints (tendons and ligaments) and the muscles around<|endoftext|>Log In

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.