INDEX

Explanations

plans or proposals for improvements or changes

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

PrototypeOf

-0.38

 someone

-0.31

 diapers

-0.28

 faltar

-0.28

 somebody

-0.27

彼が

-0.27

ReadInt

-0.26

someone

-0.25

 napi

-0.25

="@+

-0.25

POSITIVE LOGITS

 itſelf

0.74

 itself

0.73

its

0.68

 Оно

0.65

 informée

0.63

GEBURTSDATUM

0.63

astify

0.63

itself

0.62

 InputDecoration

0.59

 Itself

0.56

Activations Density 0.980%

plans or proposals for improvements or changes

No Comments

No Known Activations