word just before a 'reason is' explanation

What this neuron does

This neuron lights up on the word right before a point is explained, like 'the reason was simple' or 'for good or ill'. It seems to expect that an explanation is coming next. Turning it up made the text wander without 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

   down all of your classes, the assignments, the time and day things are due. For some men, this could be very helpful. Planners exist for a reason–they actually work. Maybe it'll help you be more responsible in school.⏎

2. #2 peak

   seen as newsworthy.⏎But the closure of the Samchully factory attracted the attention of the major media. The reason was simple: the factory was the last in Seoul to produce coal briquettes, known

3. #3 peak

   The latter can be defended, but the former demands us to jettison its nationalistic undertones, and implicitly acknowledges the complexities of nationalism that drives societies for good or ill.⏎Nationalism becomes

4. #4 peak

   are essential for keeping millions of women and families above the poverty line. We know that women disproportionately rely on these programs, making the message from the state poverty data clear: Programs for low-income families and individuals must be protected. We sincerely hope Congress

5. #5 peak

   ) tells us what the audience will see and hear, starting with the beginning<|endoftext|>Unearthing Ancient History in Tuscany⏎The goal of this expedition is simple: to help experts reconstruct the complex past of Populonia as fully as possible

6. #6 peak

   clearly accelerating rapidly – from a crawl in biblical times, to a jog in the 20th century, to a sprint now.⏎The reason for this is simple. The more technology you have at your disposal, the easier it is to invent

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

   of Superfund Sites in the United States. As part of NIEHS's Superfund Research Program, the Puerto Rico Testsite for Exploring Contamination Threats (PROTECT) is exploring the relationships

2. #2 peak

   War, either through their own family history, links to their local community or because of its long-term impact on the world we live in today. Through projects like Faces of the First World War and the

3. #3 peak

   public to help find out more about all the servicemen whose images make up the online exhibition but about whom very little else is known.⏎The collection of photographs - Faces of the First World War - is part of the IWM's preparations to mark

4. #4 peak

   side Congress and the CFPB, the Native American Financial Services Association (NAFSA) is dedicated to providing financial literacy to everyone. Through their Financial Literacy Program, NAFSA provides free financial literacy modules and tools with easy-to-under

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.