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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

venue

-0.74

yon

-0.74

Û

-0.74

zone

-0.70

idon

-0.69

coal

-0.68

band

-0.68

 Auction

-0.67

bands

-0.67

endez

-0.67

POSITIVE LOGITS

¥ŀ

0.71

quished

0.69

Jag

0.65

ĪĴ

0.63

 carbohyd

0.63

Ti

0.61

illus

0.61

ascal

0.60

 absorbed

0.59

 sustained

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.