INDEX

Explanations

the word "ang."

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

 cref

-1.54

 ...)

-1.41

 orderly

-1.39

...?"

-1.39

havior

-1.38

curs

-1.36

 Nurs

-1.34

uminous

-1.32

../../

-1.31

cens

-1.30

POSITIVE LOGITS

recht

2.14

amycin

1.81

ue

1.70

ulation

1.67

ely

1.66

istically

1.64

istical

1.63

aku

1.63

kok

1.63

aea

1.61

Activations Density 0.037%

the word "ang."

No Comments

No Known Activations