INDEX

Explanations

writing web scripts

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

code or scripting contexts, particularly methods that output or write content (like `document.write`, `Response.Write`, or script object creation).

oai_token-act-pair · claude-4-5-sonnet Triggered by @bfgcn7frf5

The neuron fires on programming‐language code snippets—especially script function calls like document.write, document.writeln, and other scripting keywords.

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

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_10/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

’

-1.74

-1.73

↵↵

-1.52

原本

-1.47

 सभी

-1.42

or

-1.41

 для

-1.40

犇

-1.36

we

-1.36

 hjælp

-1.31

POSITIVE LOGITS

}))

1.63

亹

1.59

1.52

甃

1.52

specifically

1.52

within

1.48

翃

1.48

both

1.46

鬭

1.46

 genere

1.45

Activations Density 0.010%

writing web scripts

code or scripting contexts, particularly methods that output or write content (like `document.write`, `Response.Write`, or script object creation).

The neuron fires on programming‐language code snippets—especially script function calls like document.write, document.writeln, and other scripting keywords.

No Comments

No Known Activations

writing web scripts

code or scripting contexts, particularly methods that output or write content (like `document.write`, `Response.Write`, or script object creation).

The neuron fires on programming‐language code snippets—especially script function calls like document.write, document.writeln, and other scripting keywords.

No Comments

No Known Activations