INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

yrinth

-0.83

 Sinai

-0.75

 Palest

-0.70

 nomine

-0.68

 Scully

-0.68

 Canterbury

-0.67

 Cyprus

-0.66

itely

-0.65

itialized

-0.64

 Guam

-0.64

POSITIVE LOGITS

puff

0.81

jack

0.72

portation

0.67

ãģ®å®

0.65

FTA

0.65

mage

0.64

 accessory

0.64

 Flavor

0.64

favorite

0.63

bows

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.