INDEX

Explanations

mentions of governmental entities, specifically ministries

oai_token-act-pair · gpt-3.5-turbo

references to government ministries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 taco

-0.73

âľ

-0.72

omet

-0.69

 brow

-0.65

 Angelo

-0.64

uckle

-0.63

ander

-0.63

³

-0.63

 characterize

-0.62

kins

-0.62

POSITIVE LOGITS

 Ministry

3.78

 ministry

2.82

 ministries

2.31

 Minister

2.15

 Ministers

1.82

 minister

1.71

minist

1.64

 Department

1.59

 Directorate

1.54

Department

1.42

Activations Density 0.010%

mentions of governmental entities, specifically ministries

references to government ministries

No Comments

No Known Activations