INDEX

Explanations

phrases indicating the revelation or clarification of information

oai_token-act-pair · gpt-3.5-turbo

phrases about illuminating or revealing information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

icians

-0.72

heid

-0.68

ournals

-0.67

teness

-0.66

erity

-0.65

DragonMagazine

-0.63

IAN

-0.61

HCR

-0.61

DonaldTrump

-0.60

£ı

-0.60

POSITIVE LOGITS

urst

0.89

iencies

0.86

ritten

0.81

ench

0.80

itten

0.78

pload

0.77

 tears

0.77

irts

0.76

irth

0.76

onds

0.76

Activations Density 0.034%

phrases indicating the revelation or clarification of information

phrases about illuminating or revealing information

No Comments

No Known Activations