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

 sequ

-0.72

Fram

-0.64

edom

-0.61

 regress

-0.60

 Azerb

-0.60

 hypothesis

-0.60

 reflected

-0.58

 strand

-0.58

 happening

-0.57

inyl

-0.57

POSITIVE LOGITS

ulia

0.73

ibles

0.69

berra

0.68

soever

0.68

ventory

0.68

olic

0.67

 Polaris

0.67

 Volunteer

0.64

 Mile

0.64

anasia

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.