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

anding

-0.76

 disg

-0.70

 transplant

-0.70

 redist

-0.67

 whiff

-0.66

Exit

-0.66

wine

-0.66

incinn

-0.66

 mascul

-0.64

catentry

-0.64

POSITIVE LOGITS

rodu

0.74

ĨĴ

0.74

lations

0.74

rays

0.73

 others

0.69

 Christy

0.68

 Paula

0.68

reth

0.67

RAL

0.65

swers

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.