INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ipedia

-0.76

OTOS

-0.76

ORN

-0.71

vich

-0.70

oros

-0.67

living

-0.66

 seiz

-0.66

lins

-0.64

ulative

-0.63

below

-0.63

POSITIVE LOGITS

 Prob

0.70

 Marian

0.70

heit

0.69

fri

0.68

 Hazel

0.66

ogy

0.64

 Administ

0.64

Fri

0.64

 Manit

0.64

 bets

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.