INDEX

Explanations

references to transforming into different objects or characters

oai_token-act-pair · gpt-3.5-turbo

references to imaginative or absurd scenarios

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

Initial

-0.87

imester

-0.86

gression

-0.85

Recommend

-0.85

Register

-0.83

autions

-0.82

Media

-0.82

mediate

-0.82

Policy

-0.81

requent

-0.80

POSITIVE LOGITS

 dinosaurs

1.35

 pengu

1.30

 Godzilla

1.30

 dolphins

1.29

 Bigfoot

1.26

 sharks

1.23

 UFOs

1.22

 gorilla

1.22

 dinosaur

1.22

 Dracula

1.20

Activations Density 0.725%

references to transforming into different objects or characters

references to imaginative or absurd scenarios

No Comments

No Known Activations