INDEX

Explanations

references to video games and gaming-related terminology

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

ĳ

-1.90

İ

-1.68

anding

-1.63

Ĳ

-1.55

ł

-1.54

 cref

-1.53

©

-1.49

 deportation

-1.46

·

-1.43

POSITIVE LOGITS

 consoles

2.03

play

1.99

manship

1.77

 scene

1.77

bird

1.76

avia

1.73

Spot

1.70

 chang

1.69

 scenes

1.68

 starter

1.65

Activations Density 0.244%

references to video games and gaming-related terminology

No Comments

No Known Activations