INDEX

Explanations

content source attributions at the end of articles

oai_token-act-pair · gpt-3.5-turbo

components related to photo credits and captions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 disgu

-0.78

 infiltr

-0.75

oleon

-0.71

 creep

-0.69

ibaba

-0.68

 reve

-0.67

 idol

-0.67

ogly

-0.67

osaurs

-0.66

lycer

-0.66

POSITIVE LOGITS

Getty

1.31

Flickr

1.07

AFP

1.03

Bloomberg

1.00

Courtesy

1.00

EVA

1.00

Library

0.98

Film

0.93

WAR

0.92

NRS

0.90

Activations Density 0.038%

content source attributions at the end of articles

components related to photo credits and captions

No Comments

No Known Activations