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

ticket

-0.68

oddy

-0.67

ribing

-0.63

versions

-0.62

performance

-0.61

 Cassidy

-0.60

models

-0.60

performing

-0.59

 sidx

-0.57

opp

-0.57

POSITIVE LOGITS

 rede

0.78

rez

0.75

 Dame

0.72

ARY

0.69

ngth

0.68

zers

0.67

lite

0.65

alm

0.65

heit

0.64

¬¼

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.