INDEX

Explanations

proper nouns indicating a person's origin or occupation

oai_token-act-pair · gpt-3.5-turbo

phrases or references indicating the origin of individuals or groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

faced

-0.76

proven

-0.74

potion

-0.73

lethal

-0.72

ogene

-0.72

othe

-0.72

tons

-0.71

 showc

-0.70

endix

-0.70

isphere

-0.69

POSITIVE LOGITS

 afar

1.34

 abroad

1.02

 whence

0.92

Uni

0.90

 whom

0.85

 inside

0.84

 Brune

0.80

 within

0.78

 Auckland

0.78

 thence

0.78

Activations Density 0.147%

proper nouns indicating a person's origin or occupation

phrases or references indicating the origin of individuals or groups

No Comments

No Known Activations