INDEX

Explanations

references to distance and proximity

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

 tiennent

-0.41

 secretos

-0.40

Тарихы

-0.40

 sorpresas

-0.36

estión

-0.36

 gestionar

-0.35

 consigo

-0.35

worte

-0.35

 surpresa

-0.34

toEqual

-0.34

POSITIVE LOGITS

Datuak

0.66

EndContext

0.66

 lenker

0.52

 yonder

0.52

StoryboardSegue

0.52

SourceChecksum

0.50

richTextPanel

0.48

OGND

0.48

 ujednoznacz

0.48

 estekak

0.48

Activations Density 0.034%

references to distance and proximity

No Comments

No Known Activations