INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 lapt

-0.79

 fragrance

-0.73

BuyableInstoreAndOnline

-0.73

ãĥī

-0.71

icion

-0.69

inav

-0.69

ãĥ¡

-0.67

VW

-0.66

nir

-0.66

lux

-0.65

POSITIVE LOGITS

annels

0.72

gee

0.71

afety

0.67

hide

0.67

advertising

0.67

ports

0.66

ridge

0.66

wcs

0.64

address

0.64

ickey

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.