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

iers

-0.82

erer

-0.66

idas

-0.65

jet

-0.64

 iCloud

-0.64

ieri

-0.64

 (Â£

-0.64

oms

-0.63

 suscept

-0.62

vo

-0.62

POSITIVE LOGITS

Ŀ

1.57

¬¼

0.86

 Cosponsors

0.76

soDeliveryDate

0.74

%]

0.74

ãĥĺ

0.72

ablishment

0.71

Dex

0.70

ãĤ¼ãĤ¦ãĤ¹

0.68

 Referred

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.