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

 Stre

-0.68

 Planetary

-0.65

 Radar

-0.63

 Crusader

-0.63

 Underworld

-0.62

ãĥĭ

-0.60

 Struggle

-0.59

RGB

-0.58

 Survivor

-0.58

nesses

-0.57

POSITIVE LOGITS

eday

0.83

irrel

0.83

rency

0.79

advertisement

0.75

hid

0.74

SPONSORED

0.73

rea

0.71

DragonMagazine

0.71

apy

0.70

rex

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.