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

 Flavoring

-0.72

 Dill

-0.72

haw

-0.67

 Invention

-0.64

wash

-0.63

 Bung

-0.63

çİĭ

-0.62

 BaseType

-0.60

packing

-0.60

mop

-0.58

POSITIVE LOGITS

rences

0.80

atars

0.69

 yours

0.66

FAULT

0.64

eries

0.60

 adversity

0.60

 Scorp

0.60

ucket

0.60

 linem

0.59

foundland

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.