INDEX

Explanations

relationships and comparisons within data sets or entities

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

Edd

-0.43

hematical

-0.43

 priori

-0.42

prudence

-0.42

tij

-0.41

 burlesque

-0.40

 reven

-0.39

 Mlle

-0.38

ghum

-0.38

 genomes

-0.38

POSITIVE LOGITS

 protoimpl

0.70

ThroughAttribute

0.50

 &___

0.47

jface

0.46

 defaultstate

0.45

 المعيارى

0.45

UnusedPrivate

0.42

Atsauces

0.42

bootstrapcdn

0.41

]")]

0.40

Activations Density 0.286%

relationships and comparisons within data sets or entities

No Comments

No Known Activations