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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ensual

-0.71

xit

-0.70

 Shutterstock

-0.68

 nodd

-0.68

luster

-0.68

jri

-0.66

lesiastical

-0.65

gling

-0.62

 Goods

-0.62

ja

-0.61

POSITIVE LOGITS

othal

0.67

ARC

0.67

cells

0.66

 enable

0.65

etts

0.65

actionDate

0.63

 alignment

0.63

 squads

0.62

ãĥķ

0.61

isers

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.