INDEX

Explanations

copyright notices or file descriptions

np_acts-logits-general · gemini-2.5-flash-lite

The neuron strongly activates on the major keywords and labels in source‐file headers (e.g. license/copyright boilerplate and top‐level declarations like namespace), essentially picking out the header metadata section of code files.

oai_token-act-pair · o4-mini Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_34/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 darum

-0.83

 مساء

-0.75

LayoutParams

-0.75

 τους

-0.75

詣

-0.73

)}}

-0.70

يزات

-0.70

/**

-0.70

oter

-0.69

 mitään

-0.69

POSITIVE LOGITS

↑

0.90

on

0.84

 maksi

0.84

Elke

0.83

か

0.82

 ли

0.82

 committing

0.79

 اخبار

0.78

Archiv

0.77

Morgen

0.77

Activations Density 0.001%

copyright notices or file descriptions

The neuron strongly activates on the major keywords and labels in source‐file headers (e.g. license/copyright boilerplate and top‐level declarations like namespace), essentially picking out the header metadata section of code files.

No Comments

No Known Activations

copyright notices or file descriptions

The neuron strongly activates on the major keywords and labels in source‐file headers (e.g. license/copyright boilerplate and top‐level declarations like namespace), essentially picking out the header metadata section of code files.

No Comments

No Known Activations

copyright notices or file descriptions

The neuron strongly activates on the major keywords and labels in source‐file headers (e.g. license/​copyright boilerplate and top‐level declarations like namespace), essentially picking out the header metadata section of code files.

No Comments

No Known Activations

copyright notices or file descriptions

The neuron strongly activates on the major keywords and labels in source‐file headers (e.g. license/​copyright boilerplate and top‐level declarations like namespace), essentially picking out the header metadata section of code files.

No Comments

No Known Activations

The neuron strongly activates on the major keywords and labels in source‐file headers (e.g. license/copyright boilerplate and top‐level declarations like namespace), essentially picking out the header metadata section of code files.

The neuron strongly activates on the major keywords and labels in source‐file headers (e.g. license/copyright boilerplate and top‐level declarations like namespace), essentially picking out the header metadata section of code files.