INDEX

Explanations

terms related to rankings and top selections in various contexts

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

ranking or top lists, especially in statistical or academic contexts.

oai_token-act-pair · claude-3-5-haiku-20241022 Triggered by @neilrathi

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

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

 StatelessWidget

-0.50

Teb

-0.43

UnusedPrivate

-0.41

yatı

-0.40

CloseOperation

-0.40

 StringField

-0.39

 CGSize

-0.39

aarrggbb

-0.38

bearded

-0.38

HasAnnotation

-0.38

POSITIVE LOGITS

top

0.70

mejores

0.59

TOP

0.57

getTop

0.57

 ranking

0.57

出版年

0.56

Top

0.56

Top

0.56

 rankings

0.56

 ranked

0.56

Activations Density 0.158%

terms related to rankings and top selections in various contexts

ranking or top lists, especially in statistical or academic contexts.

No Comments

No Known Activations