INDEX

Explanations

references to photo credits in written content

oai_token-act-pair · gpt-3.5-turbo

references to image credits or sources

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ormal

-0.82

ced

-0.77

etime

-0.77

cible

-0.76

ggle

-0.73

 disarm

-0.72

stood

-0.70

etheless

-0.68

soType

-0.67

 partic

-0.67

POSITIVE LOGITS

 Images

1.26

Images

0.96

 IMAGES

0.93

 Thumbnails

0.92

 Library

0.88

 caption

0.87

 Libraries

0.87

sburg

0.87

 Painting

0.86

Library

0.86

Activations Density 0.018%

references to photo credits in written content

references to image credits or sources

No Comments

No Known Activations