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

 millenn

-0.88

 mosqu

-0.87

 enthusi

-0.86

 tiss

-0.81

 Palestin

-0.80

 indo

-0.79

 gobl

-0.70

 confir

-0.70

 Phar

-0.69

nav

-0.69

POSITIVE LOGITS

 Solitaire

0.65

DonaldTrump

0.64

 ..........

0.64

avering

0.63

-----------

0.62

icken

0.61

icides

0.60

arenthood

0.60

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

0.59

SPONSORED

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.