INDEX

Explanations

references to the United States

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 complete

-1.96

 true

-1.88

 mean

-1.86

 immediate

-1.64

respectively

-1.57

awning

-1.56

 truly

-1.56

 late

-1.50

 empty

-1.50

 normal

-1.49

POSITIVE LOGITS

cript

2.23

ÅĽci

1.88

creen

1.86

enos

1.86

 bases

1.73

ygen

1.71

iÃ¨re

1.66

idelines

1.65

 arsenal

1.65

volt

1.65

Activations Density 4.841%

references to the United States

No Comments

No Known Activations