INDEX

Explanations

phrases related to legal definitions and consequences of actions

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

arynge

-1.91

lication

-1.77

because

-1.70

Ð³Ð¾

-1.68

reement

-1.66

suffix

-1.66

falls

-1.64

procedure

-1.56

arlier

-1.51

arÃ©

-1.50

POSITIVE LOGITS

ħ

2.10

 remotely

1.76

ĭ

1.73

ļ

1.61

ĵ

1.56

·¸

1.53

¾

1.53

ĸ´

1.50

 Evan

1.48

 access

1.48

Activations Density 2.512%

phrases related to legal definitions and consequences of actions

No Comments

No Known Activations