INDEX

Explanations

images or photographs posted on social media

oai_token-act-pair · gpt-3.5-turbo

references to images and photographs shared on social media

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Construction

-0.68

asions

-0.62

BUS

-0.62

orney

-0.61

 Nadu

-0.60

ivas

-0.58

osponsors

-0.57

anship

-0.57

utical

-0.57

byter

-0.56

POSITIVE LOGITS

of

1.15

 depicting

1.04

 showing

1.01

 thereof

0.97

 purportedly

0.95

 resembling

0.86

 featuring

0.85

 illustrating

0.85

 mont

0.83

 purported

0.81

Activations Density 0.102%

images or photographs posted on social media

references to images and photographs shared on social media

No Comments

No Known Activations