INDEX

Explanations

npm update, .dimensions, define

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

code or technical documentation, particularly formatted lists with delimiters like equals signs, hyphens, or pound signs.

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

The neuron fires on numeric tokens—any standalone digits or multi-digit number (including version numbers, IDs, and constants) in the text.

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

着一个

-1.98

phòng

-1.62

:'/

-1.61

着自己的

-1.60

了一会儿

-1.57

着一

-1.52

これから

-1.45

toString

-1.45

-1.42

 tqdm

-1.41

POSITIVE LOGITS

2.50

allez

1.65

 were

1.53

ませんでした

1.52

guisement

1.46

 only

1.45

).

1.45

the

1.42

 mịn

1.42

shein

1.41

Activations Density 1.544%

npm update, .dimensions, define

code or technical documentation, particularly formatted lists with delimiters like equals signs, hyphens, or pound signs.

The neuron fires on numeric tokens—any standalone digits or multi-digit number (including version numbers, IDs, and constants) in the text.

No Comments

No Known Activations

npm update, .dimensions, define

code or technical documentation, particularly formatted lists with delimiters like equals signs, hyphens, or pound signs.

The neuron fires on numeric tokens—any standalone digits or multi-digit number (including version numbers, IDs, and constants) in the text.

No Comments

No Known Activations