INDEX

Explanations

keywords related to restrictions or limitations

oai_token-act-pair · gpt-3.5-turbo

phrases indicating limitations or prohibitions

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

adle

-0.78

buster

-0.73

 LOVE

-0.73

angel

-0.73

liest

-0.71

oes

-0.71

tale

-0.70

iron

-0.69

prep

-0.69

wives

-0.68

POSITIVE LOGITS

 restricted

3.72

restricted

2.54

 unrestricted

2.14

ricted

2.09

 limited

1.88

 restrict

1.80

 confined

1.80

 restriction

1.77

 constrained

1.76

 Restrict

1.73

Activations Density 0.013%

keywords related to restrictions or limitations

phrases indicating limitations or prohibitions

No Comments

No Known Activations