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

HI

-0.80

uci

-0.75

onge

-0.70

ake

-0.68

ilk

-0.67

 TYPE

-0.66

AH

-0.64

uo

-0.64

athed

-0.63

ggles

-0.63

POSITIVE LOGITS

rax

0.79

 Siberian

0.69

 Tenth

0.68

 Zeal

0.65

ãĤ¶

0.65

áµ

0.65

icum

0.63

 symb

0.63

 Operator

0.63

 Prism

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.