INDEX

Explanations

references to external sources or attributions

oai_token-act-pair · gpt-3.5-turbo

references to the source or attribution of information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

inary

-0.93

eared

-0.77

ecause

-0.76

earing

-0.74

 footing

-0.73

 teeth

-0.72

ines

-0.71

 frogs

-0.70

bats

-0.70

iggs

-0.70

POSITIVE LOGITS

 Wikimedia

1.03

 Shutterstock

0.92

 REUTERS

0.87

 Sina

0.86

Yon

0.83

 Yelp

0.82

 Everett

0.82

 Huffington

0.81

Via

0.80

WS

0.80

Activations Density 0.017%

references to external sources or attributions

references to the source or attribution of information

No Comments

No Known Activations