INDEX

Explanations

references to case numbers or legal citations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

uated

-1.61

roc

-1.60

lasses

-1.57

ually

-1.55

ded

-1.49

its

-1.47

gered

-1.45

acles

-1.43

acular

-1.42

ributed

-1.41

POSITIVE LOGITS

ir

1.76

zzle

1.73

ve

1.55

arl

1.51

 kidding

1.50

utum

1.48

zzles

1.46

jon

1.46

qual

1.46

ament

1.44

Activations Density 0.093%

references to case numbers or legal citations

No Comments

No Known Activations