INDEX

Explanations

phrases related to documents, articles, or written pieces

oai_token-act-pair · gpt-3.5-turbo

symbols and punctuations commonly associated with formatting or markup language

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

eur

-0.61

 manif

-0.55

ppard

-0.54

foc

-0.52

ific

-0.52

itiz

-0.51

ifiable

-0.50

 graft

-0.50

 diseng

-0.50

ier

-0.50

POSITIVE LOGITS

 Expand

0.83

actionDate

0.76

 Advertisement

0.76

 Provided

0.74

Reviewer

0.74

ĸļ

0.72

 Transcript

0.72

 SHARES

0.71

BuyableInstoreAndOnline

0.70

 largeDownload

0.70

Activations Density 0.106%

phrases related to documents, articles, or written pieces

symbols and punctuations commonly associated with formatting or markup language

No Comments

No Known Activations