INDEX

Explanations

specific years mentioned in a historical context

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

¼

-1.81

Ļ

-1.74

Ħ

-1.73

ĥ

-1.73

Ĺ

-1.68

 recent

-1.58

ager

-1.57

¢

-1.57

Ģ

-1.57

Ĩ

-1.56

POSITIVE LOGITS

nai

1.64

 anniversary

1.63

 Ð³Ð¾Ð´

1.62

å¹´

1.59

iolet

1.57

RSV

1.55

ITED

1.54

itat

1.53

IRA

1.48

IGEN

1.47

Activations Density 0.056%

specific years mentioned in a historical context

No Comments

No Known Activations