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

éŃĶ

-0.65

isphere

-0.64

GROUP

-0.64

OB

-0.63

Op

-0.62

TIME

-0.61

ANN

-0.60

entious

-0.60

NAT

-0.59

 Dealer

-0.59

POSITIVE LOGITS

ammy

0.76

ertodd

0.74

 Gareth

0.74

andro

0.73

aeper

0.70

ÃŃa

0.70

ingham

0.69

DF

0.69

chuk

0.68

ault

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.