INDEX

Explanations

online presence information, such as social media profiles or blogs, related to specific individuals

oai_token-act-pair · gpt-3.5-turbo

references to professional roles and affiliations

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

 delaying

-0.70

 deduct

-0.65

 ration

-0.63

 deduction

-0.61

 extinguished

-0.60

 haul

-0.60

 laying

-0.60

 sticking

-0.60

 deterrent

-0.59

©¶æ¥µ

-0.59

POSITIVE LOGITS

Copyright

0.92

Disclaimer

0.88

ocument

0.82

Contents

0.80

Quotes

0.77

Advertisements

0.76

usha

0.76

Credits

0.75

Feel

0.75

Website

0.75

Activations Density 0.450%

online presence information, such as social media profiles or blogs, related to specific individuals

references to professional roles and affiliations

No Comments

No Known Activations