INDEX

Explanations

mentions of custom-made or customized items/products

oai_token-act-pair · gpt-3.5-turbo

references to customization and personalized elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ï¸

-0.75

loo

-0.75

Æ

-0.74

OPE

-0.72

 CrossRef

-0.70

 Kinn

-0.69

wick

-0.69

minster

-0.68

trump

-0.68

=-=-=-=-

-0.68

POSITIVE LOGITS

isable

1.60

izations

1.49

isations

1.33

ised

1.33

arily

1.29

izable

1.26

isation

1.23

ization

1.07

izing

1.02

ized

0.98

Activations Density 0.030%

mentions of custom-made or customized items/products

references to customization and personalized elements

No Comments

No Known Activations