INDEX

Explanations

acronyms or technical terms related to specific fields such as medical imaging, engineering, and law

oai_token-act-pair · gpt-3.5-turbo

abbreviations and acronyms related to scientific and technical terms

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

 Johns

-0.74

jon

-0.70

 Levine

-0.70

 Hughes

-0.68

roman

-0.67

vae

-0.67

Elk

-0.65

gor

-0.65

hern

-0.64

amar

-0.63

POSITIVE LOGITS

FORM

1.48

ING

1.34

LINE

1.32

OUS

1.32

LESS

1.32

WOR

1.31

DATA

1.30

ERS

1.29

NING

1.28

FUL

1.26

Activations Density 0.158%

acronyms or technical terms related to specific fields such as medical imaging, engineering, and law

abbreviations and acronyms related to scientific and technical terms

No Comments

No Known Activations