INDEX

Explanations

phrases related to clothing and undressing

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

ritic

-0.76

 Galile

-0.76

 Kern

-0.76

Torrent

-0.73

rencies

-0.73

BTC

-0.73

ithmetic

-0.70

Capital

-0.69

KNOWN

-0.69

 msec

-0.68

POSITIVE LOGITS

 worn

1.66

 sleeves

1.36

 gown

1.32

 wearer

1.29

 draped

1.23

cloth

1.22

 sleeve

1.22

 jacket

1.21

 clothing

1.19

 garment

1.18

Activations Density 0.282%

phrases related to clothing and undressing

No Comments

No Known Activations