INDEX

Explanations

references to legal terms and frameworks

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

AndEndTag

-0.59

ỡng

-0.54

 resourceCulture

-0.46

esgue

-0.45

gebob

-0.44

 старости

-0.42

hattan

-0.41

ificaciones

-0.40

 సౌకర్య

-0.39

BeginContext

-0.39

POSITIVE LOGITS

Tembelea

0.51

Datuak

0.48

featureID

0.47

StructEnd

0.46

Билгалдахарш

0.43

Sucesor

0.42

Diweddarwch

0.41

 FetchType

0.41

PostInfinity

0.39

Portale

0.38

Activations Density 0.561%

references to legal terms and frameworks

No Comments

No Known Activations