INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

natureconservancy

-0.79

ju

-0.79

ocking

-0.75

 è£ıç

-0.71

amy

-0.71

aez

-0.68

ocked

-0.66

 cryptoc

-0.65

andel

-0.64

NK

-0.64

POSITIVE LOGITS

 Thib

0.69

 Supporters

0.68

rek

0.64

rette

0.63

 Strait

0.62

 Guatem

0.62

tymology

0.61

 hypoc

0.61

Mou

0.59

 Purpose

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.