INDEX

Explanations

occurrences of the word "the"

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.54

’

-0.46

 Homo

-0.43

 pire

-0.43

 bore

-0.43

 nosotros

-0.42

пер

-0.41

 leva

-0.41

чо

-0.40

Ob

-0.39

POSITIVE LOGITS

InjectAttribute

1.38

 мәкал

1.34

Vidite

1.30

GEBURTSDATUM

1.20

Geplaatst

1.19

rungsseite

1.16

webElementXpaths

1.15

LookAnd

1.14

 nahilalakip

1.12

oa̍t

1.11

Activations Density 0.249%

occurrences of the word "the"

No Comments

No Known Activations