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

uria

-0.76

eme

-0.72

yles

-0.71

oos

-0.71

anwhile

-0.69

culosis

-0.66

prise

-0.65

osate

-0.65

oiler

-0.64

orf

-0.63

POSITIVE LOGITS

DEV

0.88

XT

0.82

DragonMagazine

0.79

GV

0.77

Accessory

0.73

VICE

0.68

IRC

0.66

MpServer

0.66

duino

0.65

DO

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.