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

endor

-0.70

icro

-0.61

elight

-0.61

 Mesa

-0.60

ipes

-0.59

 Lamar

-0.59

aired

-0.58

itialized

-0.58

psons

-0.57

urion

-0.56

POSITIVE LOGITS

Traditional

0.72

 Cust

0.72

cery

0.68

 Shutterstock

0.65

 absentee

0.64

jriwal

0.63

©¶æ

0.63

 Marriage

0.62

VERTISEMENT

0.61

ansom

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.