INDEX

Explanations

mentions of legal or political terms related to civil rights and civil services

oai_token-act-pair · gpt-3.5-turbo

phrases related to civil rights and social justice issues

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

 Kenny

-0.86

 Stard

-0.82

 Clyde

-0.80

 Ruff

-0.80

Wee

-0.80

 Benny

-0.77

 Sultan

-0.75

 Ratt

-0.74

 Guth

-0.74

 Sponge

-0.74

POSITIVE LOGITS

rights

1.38

sector

1.34

purpose

1.32

property

1.31

agency

1.28

estate

1.26

energy

1.25

policy

1.24

health

1.24

intelligence

1.24

Activations Density 0.103%

mentions of legal or political terms related to civil rights and civil services

phrases related to civil rights and social justice issues

No Comments

No Known Activations