INDEX

Explanations

phrases related to government entities and policies

oai_token-act-pair · gpt-3.5-turbo

references to governmental discussions or policies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

oux

-0.73

arnaev

-0.73

Scar

-0.73

bleacher

-0.72

socket

-0.72

ritch

-0.70

é¾įåĸļå£«

-0.68

Americ

-0.68

ricane

-0.68

 Cubs

-0.67

POSITIVE LOGITS

 government

1.81

 ministries

1.80

 Ministers

1.78

 Minister

1.77

 minister

1.77

Government

1.77

 ministry

1.76

 ministers

1.75

 Ministry

1.75

government

1.73

Activations Density 0.555%

phrases related to government entities and policies

references to governmental discussions or policies

No Comments

No Known Activations