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

terday

-0.82

ieves

-0.78

TOD

-0.76

ajor

-0.74

=]

-0.74

Ô

-0.71

ablishment

-0.70

enegger

-0.68

'.

-0.68

ãĥ¼ãĥ³

-0.68

POSITIVE LOGITS

natureconservancy

0.81

onite

0.71

ero

0.69

wolf

0.67

 Braun

0.65

vez

0.65

venants

0.65

pes

0.64

bara

0.64

eros

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.