INDEX

Explanations

patterns related to visual formatting or symbols used in written content

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

ۜ

-0.48

enderror

-0.46

 béisbol

-0.45

cyd

-0.44

 Grüsse

-0.44

strate

-0.43

lottes

-0.43

ActivityCompat

-0.43

UnusedPrivate

-0.42

 Deficit

-0.41

POSITIVE LOGITS

Autoritní

0.48

fillType

0.46

Predecesor

0.42

 typelib

0.41

ArrowToggle

0.40

 border

0.38

 borders

0.38

 cross

0.37

yym

0.36

 doors

0.36

Activations Density 0.433%

patterns related to visual formatting or symbols used in written content

No Comments

No Known Activations