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

elines

-0.87

othe

-0.82

wx

-0.72

chy

-0.72

 Unch

-0.71

ilon

-0.67

 âĶ

-0.67

venants

-0.66

ģĸ

-0.66

hots

-0.64

POSITIVE LOGITS

INAL

0.72

backer

0.68

zag

0.68

 Cabin

0.64

Pwr

0.63

paio

0.63

 Pair

0.61

 Madden

0.60

allery

0.60

 inacc

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.