INDEX

Explanations

references to the brand "Lego"

oai_token-act-pair · gpt-3.5-turbo

references to Lego

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

bern

-0.72

pressed

-0.71

sweet

-0.71

iping

-0.68

close

-0.64

ushes

-0.63

terness

-0.63

ews

-0.62

 breath

-0.62

 Limbaugh

-0.61

POSITIVE LOGITS

 Lego

1.18

 LEGO

1.13

 bricks

0.88

 Dimensions

0.82

 Minecraft

0.79

 Theft

0.75

 Ideas

0.73

 Motorsport

0.71

 Zelda

0.69

MpServer

0.69

Activations Density 0.007%

references to the brand "Lego"

references to Lego

No Comments

No Known Activations