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

merce

-0.79

gage

-0.73

assic

-0.73

oplan

-0.71

cig

-0.68

idia

-0.64

initions

-0.63

amina

-0.63

rar

-0.62

 relate

-0.62

POSITIVE LOGITS

 Unicorn

0.71

 Begin

0.70

Bul

0.68

sat

0.67

 Parables

0.65

apest

0.65

anca

0.65

 Anonymous

0.63

jer

0.62

 Ital

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.