INDEX

Explanations

specific words related to societal norms

oai_token-act-pair · gpt-3.5-turbo

references to societal norms and expectations

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

head

-0.73

wen

-0.72

RET

-0.66

iddler

-0.64

lees

-0.64

 Hidden

-0.64

 Died

-0.63

del

-0.63

zz

-0.63

lee

-0.63

POSITIVE LOGITS

 norms

3.65

 norm

1.97

 normative

1.66

 conventions

1.60

norm

1.59

 stereotypes

1.41

 standards

1.38

 ideals

1.38

Norm

1.37

 expectations

1.36

Activations Density 0.017%

specific words related to societal norms

references to societal norms and expectations

No Comments

No Known Activations