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

LD

-0.68

GMT

-0.68

 aside

-0.67

 Tiny

-0.67

 Conan

-0.66

Cth

-0.62

aky

-0.61

 Brist

-0.61

Lum

-0.60

 Candle

-0.60

POSITIVE LOGITS

andals

0.74

inally

0.73

isel

0.71

roman

0.69

hai

0.69

iak

0.68

ansas

0.67

DragonMagazine

0.67

itton

0.67

uez

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.