INDEX

Explanations

authors and their writings

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "writes."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

xon

-0.81

angular

-0.71

amic

-0.68

aned

-0.68

 Amph

-0.68

ordan

-0.68

Pont

-0.67

ANN

-0.66

phant

-0.66

acqu

-0.65

POSITIVE LOGITS

 writes

1.06

ettings

0.88

ynthesis

0.87

 WRITE

0.83

writer

0.82

 wrote

0.78

ystem

0.78

writ

0.74

 scrib

0.74

ctl

0.74

Activations Density 0.013%

authors and their writings

mentions of the word "writes."

No Comments

No Known Activations