INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

eals

-0.71

 flats

-0.69

ocations

-0.65

istries

-0.64

vit

-0.64

 elevation

-0.63

cius

-0.61

atri

-0.61

nav

-0.61

 vulner

-0.60

POSITIVE LOGITS

zsche

0.83

 Creed

0.78

bees

0.76

iqueness

0.72

 Warfare

0.69

pedia

0.67

ecause

0.67

 Chess

0.66

arthed

0.66

 Awakens

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.