INDEX

Explanations

first-person subjective expressions and references to personal experiences

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

 itſelf

-0.47

 šte

-0.47

ederen

-0.46

 Référence

-0.42

 esfuer

-0.42

üf

-0.41

ungalows

-0.40

abang

-0.40

 Volkes

-0.40

 års

-0.39

POSITIVE LOGITS

 Италијани

0.58

:✨

0.51

 estekak

0.48

IntoConstraints

0.46

RefNanny

0.46

 described

0.46

principalColumn

0.45

oredCriteria

0.43

Хьажоргаш

0.43

 outlined

0.42

Activations Density 0.049%

first-person subjective expressions and references to personal experiences

No Comments

No Known Activations