INDEX

Explanations

proper nouns and names, possibly related to legal cases or parties involved

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

oretic

-1.88

onical

-1.68

reland

-1.67

ixel

-1.66

etary

-1.63

etimes

-1.63

 generation

-1.62

xico

-1.58

ESULT

-1.57

sible

-1.56

POSITIVE LOGITS

ĺ

3.65

ī

3.53

Ń

3.51

ĳ

3.43

ı

3.38

§

3.37

½

3.35

·

3.32

ľ

3.27

Ĭ

3.26

Activations Density 0.395%

proper nouns and names, possibly related to legal cases or parties involved

No Comments

No Known Activations