INDEX

Explanations

educational or instructional content related to activities or training

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

 Infór

-0.50

getMenuInflater

-0.49

yntaxException

-0.48

ChildScrollView

-0.47

'))

-0.45

')))

-0.45

ähteet

-0.44

 defaultstate

-0.43

makeText

-0.43

 agência

-0.43

POSITIVE LOGITS

 learning

0.70

 learn

0.68

 apprend

0.60

Aprend

0.59

 taught

0.59

 training

0.59

learning

0.58

 aprend

0.57

learn

0.57

Learn

0.57

Activations Density 0.010%

educational or instructional content related to activities or training

No Comments

No Known Activations