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

acci

-0.76

Asset

-0.75

bush

-0.70

Posts

-0.69

git

-0.68

 IMAGES

-0.67

Publisher

-0.66

ŀ

-0.66

 Producer

-0.65

|--

-0.65

POSITIVE LOGITS

abulary

0.71

quished

0.70

kinson

0.67

ieving

0.67

 Corinthians

0.65

nce

0.64

irgin

0.63

 paved

0.63

lu

0.61

 Padres

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.