INDEX

Explanations

instances of the word "administration."

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

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__mlp-sm_processed/0-mlp-sm

Prompts (Dashboard)

32,768 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

torch.float32

Hook Name

blocks.0.hook_mlp_out

Hook Layer

Architecture

standard

Context Size

128

Dataset

EleutherAI/the_pile_deduplicated

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

EVER

-1.75

USY

-1.62

ubicin

-1.60

aucoup

-1.60

>&

-1.60

opol

-1.58

sed

-1.55

 Cities

-1.55

ciparum

-1.55

jours

-1.52

POSITIVE LOGITS

cript

1.71

 manual

1.69

 schedules

1.64

isters

1.63

 apparatus

1.62

 function

1.62

ister

1.55

yrinth

1.55

 process

1.54

 scene

1.54

Activations Density 0.128%

instances of the word "administration."

No Comments

No Known Activations