INDEX

Explanations

1

np_max-act · gemini-2.0-flash

This neuron responds to numeric tokens used as superscript reference or affiliation markers (e.g. the little raised numbers in author lists and citations).

oai_token-act-pair · o4-mini Triggered by @xinyanhu8

New Auto-Interp

Configuration

andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.11.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 getImage

-0.07

 NSInteger

-0.06

 Rolex

-0.06

 Lore

-0.06

_changed

-0.06

 baktı

-0.06

Park

-0.06

/she

-0.06

чает

-0.06

POSITIVE LOGITS

 spokesman

0.07

continent

0.07

(size

0.07

lease

0.07

 slowing

0.06

settings

0.06

venile

0.06

 TOUCH

0.06

 spokesperson

0.06

.Sqrt

0.06

Activations Density 0.004%

1

This neuron responds to numeric tokens used as superscript reference or affiliation markers (e.g. the little raised numbers in author lists and citations).

No Comments

No Known Activations