INDEX

Explanations

references to geographical locations, particularly the Americas and the Northern Hemisphere

oai_token-act-pair · gpt-3.5-turbo

references to the Americas and geographical contexts related to it

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

enegger

-1.07

ERAL

-0.97

agher

-0.90

alore

-0.81

raq

-0.79

CHAT

-0.77

deen

-0.77

adel

-0.77

ancies

-0.76

ements

-0.75

POSITIVE LOGITS

 Inquisition

0.87

idad

0.76

tam

0.75

 Americas

0.72

 Guerrero

0.71

 enslaved

0.70

 Scrolls

0.70

 Maya

0.69

 Haiti

0.69

 indigenous

0.68

Activations Density 0.024%

references to geographical locations, particularly the Americas and the Northern Hemisphere

references to the Americas and geographical contexts related to it

No Comments

No Known Activations