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

ftime

-0.82

ynchron

-0.68

 genomes

-0.66

pora

-0.66

rences

-0.65

ign

-0.62

ather

-0.62

athering

-0.59

gemony

-0.59

 persisted

-0.59

POSITIVE LOGITS

ebus

0.75

-+-+

0.73

erous

0.71

Hos

0.69

oxide

0.67

osure

0.65

anism

0.64

zsche

0.63

lisher

0.62

terior

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.