INDEX

Explanations

words related to watching video content online

oai_token-act-pair · gpt-3.5-turbo

content related to watching videos and online interaction prompts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 challeng

-0.80

oun

-0.76

awa

-0.73

 enthusi

-0.71

 occas

-0.68

 nodd

-0.67

 warr

-0.66

 reluct

-0.65

 comr

-0.63

 advoc

-0.63

POSITIVE LOGITS

prev

0.66

thumbnails

0.61

...]

0.57

SourceFile

0.56

 BELOW

0.56

Detroit

0.56

uers

0.54

 ÃĹ

0.52

CLOSE

0.52

}}

0.51

Activations Density 0.025%

words related to watching video content online

content related to watching videos and online interaction prompts

No Comments

No Known Activations