INDEX

Explanations

words related to the release and unveiling of products and announcements

oai_token-act-pair · gpt-3.5-turbo

phrases related to product launches and releases

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

MpServer

-0.75

CNN

-0.74

Reviewer

-0.74

zsche

-0.70

chan

-0.69

Merit

-0.68

matter

-0.66

uncture

-0.65

"]=>

-0.65

NJ

-0.65

POSITIVE LOGITS

its

0.89

 teaser

0.85

new

0.83

 redesigned

0.82

 prototypes

0.82

 updates

0.81

 their

0.77

 standalone

0.76

 versions

0.74

 refreshed

0.74

Activations Density 0.238%

words related to the release and unveiling of products and announcements

phrases related to product launches and releases

No Comments

No Known Activations