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

 Bucc

-0.79

 abst

-0.70

Reply

-0.68

wcs

-0.67

aeda

-0.67

XY

-0.65

 Typhoon

-0.64

wreck

-0.63

 defect

-0.62

ctors

-0.62

POSITIVE LOGITS

Shar

0.68

arger

0.68

agos

0.66

oker

0.65

otropic

0.63

abeth

0.62

 Shir

0.60

ett

0.60

org

0.60

isse

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.