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

RD

-0.73

 DEAD

-0.72

REG

-0.69

udeb

-0.64

UNCH

-0.64

paces

-0.64

####

-0.62

ctuary

-0.62

 Patton

-0.61

 Manson

-0.61

POSITIVE LOGITS

£ı

0.76

 Flavoring

0.72

artisan

0.67

ļé

0.67

 lobe

0.67

hospital

0.65

makers

0.65

ilies

0.64

ilian

0.64

 nuances

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.