INDEX

Explanations

mentions of specific types of written content or publications

oai_token-act-pair · gpt-3.5-turbo

mentions of articles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

pter

-0.77

asio

-0.74

 stewards

-0.71

gm

-0.70

eez

-0.69

rolet

-0.69

ris

-0.66

inav

-0.66

cffffcc

-0.65

ibr

-0.65

POSITIVE LOGITS

 Articles

1.50

 articles

1.05

Article

0.85

uggest

0.85

ettings

0.83

 âĶľ

0.81

linger

0.80

 ARTICLE

0.80

 Downloads

0.79

uggets

0.78

Activations Density 0.013%

mentions of specific types of written content or publications

mentions of articles

No Comments

No Known Activations