INDEX

Explanations

geographical and location-based terms

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

évaluateur

-0.57

GEBURTSDATUM

-0.56

findpost

-0.56

MessageTagHelper

-0.54

PerformLayout

-0.51

astify

-0.51

PreExecute

-0.50

Дереккөздер

-0.49

 AssemblyCulture

-0.49

IntoConstraints

-0.48

POSITIVE LOGITS

AndEndTag

0.45

 surla

0.40

الحياه

0.40

abilia

0.37

EndContext

0.34

 Neighborhood

0.33

дарь

0.33

 contenders

0.32

 bior

0.32

 contender

0.32

Activations Density 0.022%

geographical and location-based terms

No Comments

No Known Activations