INDEX

Explanations

specific formatting and structural elements often associated with academic or technical documents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

phobia

-0.44

gans

-0.41

harf

-0.41

sified

-0.40

enni

-0.40

meisters

-0.40

eable

-0.40

onos

-0.38

 Mahl

-0.38

 Вікіпе

-0.38

POSITIVE LOGITS

 كومونز

0.71

 betweenstory

0.57

Tikang

0.55

endpush

0.54

StructEnd

0.53

 Мексичка

0.51

OGND

0.50

 nahilalakip

0.50

 snippetHide

0.50

hyrchwyd

0.49

Activations Density 0.468%

specific formatting and structural elements often associated with academic or technical documents

No Comments

No Known Activations