INDEX

Explanations

code block punctuation

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

quotation marks and string delimiters in code and structured text formats.

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

This neuron activates on double‐quote characters marking the boundaries of string literals.

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

ᥬ

-1.52

 stumbled

-1.48

Här

-1.47

Väl

-1.45

ників

-1.44

܆

-1.43

mbros

-1.42

因為

-1.41

 about

-1.41

liese

-1.40

POSITIVE LOGITS

tlement

1.51

↵↵↵↵↵↵↵↵↵↵↵↵↵

1.49

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

1.41

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

1.38

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

1.37

只剩

1.36

↵↵↵↵↵↵↵↵↵↵↵

1.35

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

1.33

↵↵↵↵↵↵↵

1.31

↵↵↵↵↵↵↵↵↵↵↵↵↵↵

1.31

Activations Density 0.006%

code block punctuation

quotation marks and string delimiters in code and structured text formats.

This neuron activates on double‐quote characters marking the boundaries of string literals.

No Comments

No Known Activations

code block punctuation

quotation marks and string delimiters in code and structured text formats.

This neuron activates on double‐quote characters marking the boundaries of string literals.

No Comments

No Known Activations