INDEX

Explanations

programming-related terms and concepts associated with modules and libraries

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

?’

-1.80

.’”

-1.72

?”

-1.68

.’

-1.61

‘

-1.59

âĢ¦”

-1.55

!’

-1.49

!”

-1.43

 fever

-1.39

 retarded

-1.37

POSITIVE LOGITS

"}](#

1.92

*/,

1.69

1.62

lickr

1.53

/#

1.50

/.

1.50

iddell

1.47

yscall

1.47

↵

1.47

Activations Density 1.027%

programming-related terms and concepts associated with modules and libraries

No Comments

No Known Activations