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

doms

-1.10

etheless

-0.83

skirts

-0.73

 Reck

-0.70

usalem

-0.70

anmar

-0.70

 brood

-0.69

_>

-0.67

ament

-0.66

idav

-0.65

POSITIVE LOGITS

HD

1.16

onge

0.77

aila

0.76

ascar

0.71

DI

0.68

acker

0.65

iri

0.63

ARP

0.63

ais

0.61

otal

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.