INDEX

Explanations

code-related structures and method calls with specific syntax elements

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

 occurs

-1.60

ity

-1.56

 preponderance

-1.56

uced

-1.53

 routine

-1.48

 MERCHANTABILITY

-1.48

ausing

-1.46

 surface

-1.43

 grades

-1.37

inde

-1.35

POSITIVE LOGITS

microsoft

1.57

mart

1.53

clair

1.35

fony

1.32

catalina

1.32

negot

1.30

 onboard

1.30

business

1.27

 Shin

1.26

flix

1.26

Activations Density 0.648%

code-related structures and method calls with specific syntax elements

No Comments

No Known Activations