INDEX

Explanations

patent references and related formatting elements

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

ckså

-0.64

出版年

-0.53

 esternos

-0.52

 scolaire

-0.51

 quæ

-0.48

 decorados

-0.47

 chrétien

-0.46

 virtuales

-0.45

 Inſ

-0.45

 magnétique

-0.45

POSITIVE LOGITS

 utafitiHapana

0.51

 Taktlose

0.46

SequentialGroup

0.45

 endwhile

0.42

kháu

0.41

愷

0.40

eqn

0.40

PathParam

0.40

deb

0.39

 Exactos

0.38

Activations Density 0.194%

patent references and related formatting elements

No Comments

No Known Activations