INDEX

Explanations

application context code

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

the pipe character ("|") used as a delimiter or separator in structured data formats and code.

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

The neuron fires on the very first “label” or identifier token that begins a new code, link, or section—i.e. the initial word in a block (framework/class names, link labels, etc.).

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

-1.58

Sziasztok

-1.55

这让

-1.46

%");

-1.39

)");

-1.38

垕

-1.36

 rosario

-1.34

Τι

-1.32

 Поэтому

-1.30

 процессе

-1.30

POSITIVE LOGITS

了他

1.66

 őket

1.46

we

1.46

ícias

1.44

vener

1.39

 singola

1.34

 asociada

1.34

跟他

1.33

 hacerlo

1.32

 vimos

1.31

Activations Density 0.014%

application context code

the pipe character ("|") used as a delimiter or separator in structured data formats and code.

The neuron fires on the very first “label” or identifier token that begins a new code, link, or section—i.e. the initial word in a block (framework/class names, link labels, etc.).

No Comments

No Known Activations

application context code

the pipe character ("|") used as a delimiter or separator in structured data formats and code.

The neuron fires on the very first “label” or identifier token that begins a new code, link, or section—i.e. the initial word in a block (framework/class names, link labels, etc.).

No Comments

No Known Activations