INDEX

Explanations

references to educating or informing the public about various topics

oai_token-act-pair · gpt-3.5-turbo

phrases related to educating the public and communities about various issues

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Carbuncle

-0.64

catentry

-0.58

ueller

-0.56

Pg

-0.56

bourg

-0.55

CLASSIFIED

-0.54

Cooldown

-0.54

uxe

-0.54

 territ

-0.52

ãĥ¼ãĥĨãĤ£

-0.52

POSITIVE LOGITS

 about

1.30

how

1.25

 regarding

1.23

 ABOUT

1.20

 concerning

1.14

 Regarding

1.10

about

1.07

 whether

1.06

why

1.04

 About

1.01

Activations Density 0.953%

references to educating or informing the public about various topics

phrases related to educating the public and communities about various issues

No Comments

No Known Activations