INDEX

Explanations

phrases related to a specific geographical location (likely a city or region) named Gal

oai_token-act-pair · gpt-3.5-turbo

references to the term "Gal" or variations of it

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ãģį

-0.75

âķĲâķĲ

-0.71

inition

-0.69

htt

-0.69

 misleading

-0.68

olicy

-0.66

ettings

-0.66

ãĤ´ãĥ³

-0.66

uyomi

-0.64

 nomine

-0.63

POSITIVE LOGITS

axies

1.18

izoph

0.95

Gal

0.95

pha

0.91

ib

0.90

Gal

0.89

anz

0.87

antz

0.86

gal

0.85

van

0.84

Activations Density 0.005%

phrases related to a specific geographical location (likely a city or region) named Gal

references to the term "Gal" or variations of it

No Comments

No Known Activations