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

Negative Logits

Reloaded

-0.72

anners

-0.67

ndum

-0.64

ruary

-0.64

Ïģ

-0.62

ossibly

-0.62

reon

-0.62

uddy

-0.62

apo

-0.60

Raz

-0.60

POSITIVE LOGITS

ival

0.97

otype

0.87

displayText

0.77

è¦ļéĨĴ

0.74

 Surviv

0.73

obyl

0.70

emis

0.69

soDeliveryDate

0.69

inventoryQuantity

0.67

otypes

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.