INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Beir

-0.83

 Mehran

-0.75

heit

-0.70

ariat

-0.70

anwhile

-0.66

nown

-0.66

 Romero

-0.66

idium

-0.65

 Ezek

-0.64

 Dmit

-0.64

POSITIVE LOGITS

jriwal

0.76

¿

0.71

 showc

0.71

ĵ

0.70

legram

0.70

inburgh

0.68

gered

0.67

plate

0.66

Winged

0.65

Kind

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.