INDEX

Explanations

structural elements in written works

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

the

-0.68

-0.58

@"/

-0.58

podar

-0.57

El

-0.56

initWith

-0.53

el

-0.53

-0.52

-0.51

-0.50

POSITIVE LOGITS

 itſelf

1.07

 parado

0.87

ſelves

0.85

 autorytatywna

0.80

 importantly

0.79

 Anſ

0.79

 faute

0.78

 themſelves

0.77

hamdu

0.76

 reaſon

0.76

Activations Density 0.171%

structural elements in written works

No Comments

No Known Activations