INDEX

Explanations

ZopeData or XML tags

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

commas that separate items in lists or sequences of numbers.

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

The neuron fires on the initial chunk of structured identifiers or list numbers—e.g. the “Zo” at the start of a PascalCase tag or a digit with its following comma like “3,”.

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

٣

-1.58

-1.54

-1.52

是

-1.45

 tentative

-1.42

\&

-1.41

 histórica

-1.38

 что

-1.38

section

-1.37

POSITIVE LOGITS

Gbps

1.52

 古着

1.51

 zouden

1.45

 Redewendung

1.45

 trattano

1.45

 whom

1.42



1.40

 イラストレーション

1.39

炀

1.38

 whome

1.38

Activations Density 0.000%

ZopeData or XML tags

commas that separate items in lists or sequences of numbers.

The neuron fires on the initial chunk of structured identifiers or list numbers—e.g. the “Zo” at the start of a PascalCase tag or a digit with its following comma like “3,”.

No Comments

No Known Activations

ZopeData or XML tags

commas that separate items in lists or sequences of numbers.

The neuron fires on the initial chunk of structured identifiers or list numbers—e.g. the “Zo” at the start of a PascalCase tag or a digit with its following comma like “3,”.

No Comments

No Known Activations