INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

BSD

-0.74

Links

-0.67

Disk

-0.64

joy

-0.64

 Shack

-0.64

Dri

-0.63

phy

-0.63

CVE

-0.62

Secondly

-0.61

 Crescent

-0.61

POSITIVE LOGITS

umin

0.73

arel

0.68

 respir

0.68

ettings

0.68

 brewed

0.67

 nightmares

0.67

ĨĴ

0.66

etheless

0.65

 trib

0.65

meric

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.