INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Ô

-0.82

arettes

-0.79

apses

-0.72

yrinth

-0.71

akeru

-0.68

erest

-0.68

jug

-0.67

aura

-0.65

fml

-0.64

btn

-0.64

POSITIVE LOGITS

 brig

0.76

 casualty

0.73

 alike

0.73

ifi

0.69

lain

0.64

soever

0.63

 mainland

0.62

tur

0.61

Khe

0.60

 neglected

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.