line breaks in lists and wrapped text

What this neuron does

This neuron fires on line-break characters placed in the middle of a sentence or list item, often where the text was wrapped or split oddly. Turning it up made the model write with a line break after almost every word, which looks broken.

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

   and⏎African American History -⏎This website is for African American Artists and an on-line portal⏎for both African America Artists and African American History. The website⏎contains 1000's of links and 100's of

2. #2 peak

   get the⏎"gist" of the message.⏎The Internet Acronym Server⏎This site can be used to look up the meaning of an acronym (abbreviations⏎constructed from the first letters of several words, such as ADA for⏎American Dental

3. #3 peak

   York,⏎Delaware, North Carolina, Tennessee, Arkansas, South Dakota, Nebraska,⏎Kansas, Texas, New Mexico, Montana, Alaska, and California; Canada=⏎Manitoba, Alberta, British Columbia; Mexico; Brazil; Poland; Denmark

4. #4 peak

   Cattle truck and the Sheep⏎truck change places and the Engine is back on the main line?⏎In how many ways can you fit two of these yellow triangles⏎together? Can you predict the number of ways two blue triangles can⏎be

5. #5 peak

   the picture?⏎10 space travellers are waiting to board their spaceships. There⏎are two rows of seats in the waiting room. Using the rules, where⏎are they all sitting? Can you find all the possible ways?⏎What is the

6. #6 peak

   ⏎diff output format⏎to represent changes in directory structure, and extending⏎to understand these extensions.⏎Some files are neither directories nor regular files: they are unusual⏎files like symbolic links, device special files, named pipes, and⏎

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

   Dhoti Ceremony. This is a momentous day in the Telugu⏎family's calendar, which they call "Panchalu," and they celebrate it⏎in a magnificent manner with their close friends and relatives.⏎Q2. What is

2. #2 peak

   but main⏎branches absent; dichotomous branching throughout. Axils rounded, with sides⏎symmetrically developed, with side branches at times growing almost⏎horizontally for 1 - 5 mm before bending up or curling downward

3. #3 peak

   two or more servers do not interfere⏎with each other. Usually NFS file locking is handled by the⏎lockd daemon, but at the moment there is no⏎platform that performs locking 100% reliably in every situation.⏎With this

4. #4 peak

   also describes key elements of the Java API⏎library, such as I/O, the Collections Framework, the concurrency⏎utilities, the stream library, networking, event⏎handling, AWT, and more. Three chapters provide a solid introduction⏎to

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.