INDEX

Explanations

code snippets and non-english text

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

function or method calls in programming code, particularly those involving iteration, mapping, or data retrieval operations.

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

The neuron fires on programming identifiers—especially function/macro names (often in snake_case or uppercase macros) in code.

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

 diesem

-1.92

↵↵

-1.79

炻

-1.69

 území

-1.64

尔

-1.62

</strong>

-1.55

is

-1.48

Ecotoxicity

-1.44

越高

-1.41

så

-1.38

POSITIVE LOGITS

了不少

1.78

 conhece

1.60

了几

1.53

 видит

1.52

なー

1.48

了许多

1.48

一の

1.47

了很多

1.42

 mantém

1.42

了个

1.41

Activations Density 0.008%

code snippets and non-english text

function or method calls in programming code, particularly those involving iteration, mapping, or data retrieval operations.

The neuron fires on programming identifiers—especially function/macro names (often in snake_case or uppercase macros) in code.

No Comments

No Known Activations

code snippets and non-english text

function or method calls in programming code, particularly those involving iteration, mapping, or data retrieval operations.

The neuron fires on programming identifiers—especially function/macro names (often in snake_case or uppercase macros) in code.

No Comments

No Known Activations