INDEX

Explanations

phrases related to negative judgment or prejudice towards a group of people

oai_token-act-pair · gpt-3.5-turbo

phrases related to facelessness or anonymity

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

 Kaufman

-0.65

 Vaugh

-0.60

 Lyons

-0.59

 wealthier

-0.59

 Nicarag

-0.58

 prospect

-0.58

 Weir

-0.58

 Shelby

-0.57

 inexpensive

-0.57

DeV

-0.57

POSITIVE LOGITS

âĢ

1.43

 âĢ

1.28

[/

1.26

 âĶĤ

1.05

ãĢ

1.04

 âľ

1.04

</

1.03

[/

1.03

É

1.02

âĺ

1.01

Activations Density 0.855%

phrases related to negative judgment or prejudice towards a group of people

phrases related to facelessness or anonymity

No Comments

No Known Activations