INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Sexy

-0.80

 Tinder

-0.75

 Orbit

-0.71

 semen

-0.67

 Anarchy

-0.67

Pru

-0.67

Lob

-0.67

 Hardcore

-0.65

 Magnet

-0.64

 Magicka

-0.62

POSITIVE LOGITS

zyk

0.91

zynski

0.82

herer

0.80

HER

0.80

waters

0.79

irect

0.73

ledge

0.73

chev

0.72

EVA

0.72

erity

0.71

Activations Density 0.000%

No Known Activations

This feature has no known activations.