INDEX

Explanations

emails or email-related terms

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter "e."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

visors

-0.70

SPONSORED

-0.64

wreck

-0.64

ãĥīãĥ©

-0.63

fighters

-0.62

veyard

-0.59

å§«

-0.59

inguishable

-0.59

intendent

-0.59

gow

-0.59

POSITIVE LOGITS

3.34

1.55

1.38

1.37

eb

1.32

eg

1.30

 email

1.21

ec

1.19

 electronic

1.15

es

1.14

Activations Density 0.018%

emails or email-related terms

occurrences of the letter "e."

No Comments

No Known Activations