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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

jp

-0.77

nell

-0.68

 audience

-0.67

ovie

-0.63

 buildup

-0.63

was

-0.62

mma

-0.62

ollower

-0.60

hene

-0.59

onz

-0.57

POSITIVE LOGITS

£ı

0.92

ļéĨĴ

0.80

CLASSIFIED

0.74

 reincarn

0.72

 Metatron

0.70

ĨĴ

0.69

 paran

0.67

 externalToEVAOnly

0.67

 immortality

0.66

 Genetics

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.