INDEX

Explanations

words that start with "wh," indicating questions or inquiries

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

Viitteet

-0.46

Jîn

-0.44

 Trot

-0.43

getMenuInflater

-0.42

transQ

-0.41

 وتسجيلات

-0.41

 Référence

-0.41

 pense

-0.41

 setLoading

-0.40

 burning

-0.40

POSITIVE LOGITS

wh

0.70

Wh

0.66

wh

0.65

Wh

0.65

ScopeManager

0.60

WH

0.57

WH

0.47

ViewFeatures

0.47

Whi

0.46

Whi

0.44

Activations Density 0.229%

words that start with "wh," indicating questions or inquiries

No Known Activations