INDEX

Explanations

names or terms related to specific individuals

oai_token-act-pair · gpt-3.5-turbo

instances of the word "an."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ï¸ı

-0.88

 WATCHED

-0.82

hovah

-0.80

ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ

-0.74

SOURCE

-0.74

 ACTIONS

-0.72

 guiIcon

-0.69

orsi

-0.69

eatures

-0.68

 SERV

-0.68

POSITIVE LOGITS

igans

1.39

xiety

1.10

onymous

1.10

esthetic

1.06

alyst

1.05

thus

1.04

apolis

1.00

ning

1.00

odon

0.95

cient

0.94

Activations Density 0.045%

names or terms related to specific individuals

instances of the word "an."

No Comments

No Known Activations