INDEX

Explanations

intensifiers or modifiers that amplify descriptions or feelings

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ments

-1.84

ings

-1.83

 encodes

-1.77

ier

-1.58

ies

-1.58

rait

-1.58

egen

-1.57

ment

-1.56

iest

-1.51

 thus

-1.49

POSITIVE LOGITS

ppy

1.80

bject

1.76

 young

1.71

 coincidence

1.67

arious

1.63

 kids

1.63

far

1.53

ppo

1.53

 elderly

1.53

ylan

1.52

Activations Density 0.165%

intensifiers or modifiers that amplify descriptions or feelings

No Comments

No Known Activations