INDEX

Explanations

terms related to illustrations and illustrators

oai_token-act-pair · gpt-3.5-turbo

terms related to illustrations and their creators

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

etime

-0.93

ega

-0.80

tz

-0.74

roe

-0.72

ought

-0.71

inki

-0.68

uld

-0.68

rown

-0.67

cert

-0.65

oln

-0.65

POSITIVE LOGITS

 illustration

1.32

 illustrations

1.20

 illust

1.17

 Illust

1.13

 illustrating

1.10

rator

0.97

Artist

0.96

 Caption

0.92

 illustrated

0.86

 depicting

0.86

Activations Density 0.025%

terms related to illustrations and illustrators

terms related to illustrations and their creators

No Comments

No Known Activations