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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 pont

-0.72

pport

-0.70

acea

-0.68

 resusc

-0.63

AMI

-0.63

vier

-0.62

 mang

-0.61

sels

-0.61

 mans

-0.61

 sling

-0.60

POSITIVE LOGITS

raid

0.71

agon

0.70

otion

0.68

endants

0.65

inar

0.65

endant

0.64

edd

0.63

onies

0.63

omination

0.62

yt

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.