INDEX

Explanations

references to the city of Baltimore

oai_token-act-pair · gpt-3.5-turbo

references to Baltimore, particularly in the context of sports and events

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

yip

-1.03

mble

-0.84

uador

-0.83

utterstock

-0.78

BOOK

-0.77

aeda

-0.77

pter

-0.74

utherland

-0.74

berman

-0.71

igers

-0.71

POSITIVE LOGITS

 Orioles

1.23

 Ravens

1.02

 Colts

0.92

 Wire

0.87

 Baltimore

0.84

 City

0.81

 Hopkins

0.81

 County

0.80

Baltimore

0.80

 Citiz

0.77

Activations Density 0.008%

references to the city of Baltimore

references to Baltimore, particularly in the context of sports and events

No Comments

No Known Activations