INDEX

Explanations

phrases related to questions and inquiries about various topics

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

 larmes

-0.45

 pokra

-0.43

iastes

-0.43

 setw

-0.42

 SDLK

-0.42

 ricordare

-0.41

 âmes

-0.40

etsk

-0.40

 Entscheid

-0.40

upné

-0.40

POSITIVE LOGITS

we

0.65

 they

0.62

he

0.55

you

0.55

 Filipinos

0.53

帖最后由

0.52

Bakgrunnsstoff

0.48

 Мексичка

0.48

 يتيمه

0.47

 explorers

0.47

Activations Density 0.468%

phrases related to questions and inquiries about various topics

No Comments

No Known Activations