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

hov

-0.98

lesiastical

-0.75

osi

-0.71

utenberg

-0.70

agogue

-0.69

ieties

-0.68

enza

-0.67

ozyg

-0.66

oxide

-0.65

ãĤ¡

-0.65

POSITIVE LOGITS

bidden

0.73

 Cookies

0.64

 Calories

0.61

rs

0.60

 Future

0.60

 Disapp

0.60

 punitive

0.60

 Opposition

0.60

 Chips

0.59

 Batt

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.