INDEX

Explanations

references to news publications and media sources

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

jwa

-0.53

 Popis

-0.51

ulele

-0.50

 relâche

-0.50

нциклопе

-0.49

angerie

-0.48

atchewan

-0.46

Termine

-0.45

umel

-0.43

aronder

-0.43

POSITIVE LOGITS

Datuak

0.80

endphp

0.78

 resourceCulture

0.70

KommentareTeilen

0.69

Tembelea

0.69

IntoConstraints

0.65

MemoryWarning

0.64

 interlocutor

0.63

jsxFileName

0.63

ⓘ

0.62

Activations Density 0.146%

references to news publications and media sources

No Comments

No Known Activations