INDEX

Explanations

words related to the action of logging in or login credentials

oai_token-act-pair · gpt-3.5-turbo

references to login-related activities and authentication processes

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

Jere

-0.65

relative

-0.65

Israel

-0.64

Tony

-0.63

×Ķ

-0.62

_-

-0.62

general

-0.60

ahn

-0.60

Church

-0.60

POSITIVE LOGITS

 login

1.39

 credentials

1.05

 Login

1.04

 passwords

0.94

 Username

0.92

halla

0.88

 username

0.84

 authentication

0.84

 password

0.83

login

0.83

Activations Density 0.007%

words related to the action of logging in or login credentials

references to login-related activities and authentication processes

No Comments

No Known Activations