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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ipop

-0.73

alyses

-0.72

OVA

-0.71

 guiActiveUnfocused

-0.70

imeo

-0.69

istical

-0.66

IG

-0.65

imate

-0.65

wcs

-0.65

Series

-0.65

POSITIVE LOGITS

staking

0.73

 Colt

0.73

bley

0.68

 Brist

0.65

 toile

0.63

 Tuls

0.62

 Browne

0.62

 submar

0.62

 carpet

0.62

 Antar

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.