INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

apons

-0.82

ļéĨĴ

-0.75

uberty

-0.73

uckland

-0.71

aphael

-0.70

 remission

-0.67

inos

-0.66

ippery

-0.66

luster

-0.65

maxwell

-0.62

POSITIVE LOGITS

aceous

0.90

+/-

0.69

flows

0.68

medi

0.67

Ctrl

0.67

ãĥĦ

0.66

Gar

0.66

 Strongh

0.65

haus

0.64

sonian

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.