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

TAIN

-0.83

laim

-0.77

é¾įå¥ĳå£«

-0.76

illion

-0.74

ieties

-0.74

vati

-0.73

folios

-0.72

ween

-0.71

ngth

-0.71

IFIED

-0.70

POSITIVE LOGITS

 galleries

0.74

WCS

0.68

 here

0.68

BuyableInstoreAndOnline

0.67

Lap

0.66

 Flags

0.63

Champ

0.62

hots

0.62

FML

0.61

 spoiler

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.