INDEX

Explanations

names of people and organizations, perhaps related to news and events

oai_token-act-pair · gpt-3.5-turbo

unique identifiers or names associated with individuals

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

 pollutants

-0.79

 avoidance

-0.78

 storefront

-0.78

 invention

-0.77

 paperback

-0.76

 shelters

-0.76

 circulation

-0.76

 overload

-0.76

 societies

-0.75

 formation

-0.74

POSITIVE LOGITS

Wilson

1.38

Thomas

1.27

Jr

1.25

Marie

1.24

William

1.22

Nich

1.22

ï¸ı

1.22

Connor

1.20

Brend

1.19

Ryan

1.19

Activations Density 0.219%

names of people and organizations, perhaps related to news and events

unique identifiers or names associated with individuals

No Comments

No Known Activations