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

 Barg

-0.81

ä½¿

-0.78

ufact

-0.77

ãĤ©

-0.76

 actionGroup

-0.75

keleton

-0.74

 Recession

-0.74

Î³

-0.74

OSP

-0.74

udeb

-0.74

POSITIVE LOGITS

 correctness

0.69

minster

0.69

halla

0.67

mson

0.65

pip

0.63

 appointments

0.62

 rule

0.61

 mischief

0.61

 spread

0.60

roxy

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.