INDEX

Explanations

text-related terms, specifically focusing on the word "text."

oai_token-act-pair · gpt-3.5-turbo

references to text messages and text-related content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Ern

-0.75

CVE

-0.65

 Zerg

-0.59

Ã±o

-0.59

 Tide

-0.59

Bir

-0.57

 Taste

-0.57

BLIC

-0.57

 Seat

-0.57

CVE

-0.57

POSITIVE LOGITS

ured

1.53

iles

1.31

ural

1.27

area

1.26

uring

1.24

uality

1.22

urally

1.21

ures

1.19

ually

1.17

ual

1.12

Activations Density 0.045%

text-related terms, specifically focusing on the word "text."

references to text messages and text-related content

No Comments

No Known Activations