INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 semin

-0.79

è£

-0.75

 Veter

-0.66

 Galileo

-0.66

 circ

-0.63

 vigorous

-0.62

Ruk

-0.62

Zo

-0.60

 stren

-0.60

 RELE

-0.60

POSITIVE LOGITS

uden

0.98

nown

0.86

Ã£

0.85

allas

0.80

»Ĵ

0.78

ackle

0.78

amide

0.75

ettings

0.74

onz

0.73

against

0.71

Activations Density 0.000%

No Known Activations

This feature has no known activations.