INDEX

Explanations

references to legislative or governmental actions

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

Negative Logits

atories

-1.72

clud

-1.69

cludes

-1.66

king

-1.63

cciÃ³n

-1.54

 ever

-1.48

clusions

-1.45

inary

-1.43

ocular

-1.42

 cant

-1.42

POSITIVE LOGITS

Ģ

2.05

 pylori

2.01

ĺ

1.97

ĻĤ

1.90

Ĵ

1.86

Ĥ¬

1.84

ģ

1.77

ancock

1.77

aupt

1.76

Ľ

1.73

Activations Density 0.087%

references to legislative or governmental actions

No Known Activations