INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

yi

-0.82

Yah

-0.80

die

-0.78

ema

-0.76

icter

-0.73

axis

-0.72

yah

-0.72

iate

-0.69

inous

-0.69

date

-0.68

POSITIVE LOGITS

é¾įå¥ĳå£«

0.86

GOP

0.78

Christopher

0.75

 ILCS

0.73

Hamilton

0.71

Louis

0.69

atown

0.68

Kent

0.68

Ö¼

0.68

 Seaf

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.