INDEX

Explanations

comparative phrases that indicate a difference, particularly involving the word "than."

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

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__res-sm_processed/1-res-sm

Prompts (Dashboard)

32,768 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

torch.float32

Hook Name

blocks.1.hook_resid_post

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

med

-1.75

family

-1.64

inher

-1.57

fiddle

-1.55

 sleeve

-1.48

Library

-1.46

 Navigation

-1.44

Identifier

-1.43

incorporated

-1.40

overview

-1.40

POSITIVE LOGITS

ī

1.57

ĥ½

1.55

 usual

1.55

Ļª

1.55

¼

1.51

 respects

1.50

 ours

1.49

 elsewhere

1.47

acher

1.46

thood

1.46

Activations Density 0.136%

comparative phrases that indicate a difference, particularly involving the word "than."

No Comments

No Known Activations