INDEX

Explanations

terms related to politics and government officials

oai_token-act-pair · gpt-3.5-turbo

titles and positions of individuals in a political or organizational context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 stacked

-0.74

 ridic

-0.74

 unfavorable

-0.71

 deval

-0.70

 watered

-0.70

 ponies

-0.68

 contrasting

-0.68

 disposable

-0.68

 sweeping

-0.67

 punishable

-0.66

POSITIVE LOGITS

Assistant

1.32

Senior

1.30

 Founder

1.29

Director

1.28

 Coordinator

1.26

 Editor

1.21

 Director

1.21

director

1.16

Chair

1.14

CEO

1.14

Activations Density 0.312%

terms related to politics and government officials

titles and positions of individuals in a political or organizational context

No Comments

No Known Activations