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

jab

-0.73

rigan

-0.67

aughs

-0.67

href

-0.67

NRS

-0.67

odka

-0.67

pun

-0.66

à©

-0.66

mail

-0.66

agn

-0.65

POSITIVE LOGITS

ħĭ

0.78

xus

0.73

 attm

0.71

RTX

0.71

 Decay

0.70

 Planes

0.69

itialized

0.69

Ħ¢

0.69

 authenticated

0.69

arthed

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.