INDEX

Explanations

content related to templates in a programming or markup context

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

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__res-sm_processed/2-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.2.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

ĥ

-3.10

Ĥ¬

-3.04

ĥ½

-2.87

ľĵ

-2.83

Īĺ

-2.78

§

-2.77

²

-2.61

¨

-2.53

Ī

-2.49

Ĳ

-2.48

POSITIVE LOGITS

maker

1.80

kit

1.63

makers

1.52

sight

1.47

zilla

1.41

cake

1.39

regnancy

1.37

belt

1.37

noreply

1.36

Activations Density 0.179%

content related to templates in a programming or markup context

No Comments

No Known Activations