INDEX

Explanations

phrases related to something being artificially created or influenced by humans

oai_token-act-pair · gpt-3.5-turbo

phrases related to man-made phenomena or products

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ipher

-0.72

LAN

-0.67

IRE

-0.64

VERTISEMENT

-0.64

 DRAG

-0.63

 chill

-0.63

 prick

-0.63

 ILCS

-0.63

apters

-0.62

externalActionCode

-0.62

POSITIVE LOGITS

cham

0.92

roid

0.86

Fake

0.81

netflix

0.79

iful

0.74

Made

0.72

Lua

0.70

gotten

0.69

ModLoader

0.68

bye

0.68

Activations Density 0.022%

phrases related to something being artificially created or influenced by humans

phrases related to man-made phenomena or products

No Comments

No Known Activations