INDEX

Explanations

specific locations such as polling places, schools, neighborhoods, churches, and suburbs

oai_token-act-pair · gpt-3.5-turbo

locations and places mentioned in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 everywhere

-0.74

 Tasmania

-0.73

 Scotland

-0.71

 Bihar

-0.70

 Countries

-0.69

 Gujarat

-0.68

 domestically

-0.67

 Singapore

-0.67

 Mumbai

-0.66

 Seoul

-0.65

POSITIVE LOGITS

ench

0.69

SIGN

0.67

 specializing

0.63

onductor

0.62

é¾įå¥ĳå£«

0.60

mone

0.60

ļéĨĴ

0.59

SEE

0.58

istries

0.58

isky

0.58

Activations Density 0.575%

specific locations such as polling places, schools, neighborhoods, churches, and suburbs

locations and places mentioned in the text

No Comments

No Known Activations