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

theless

-0.93

terday

-0.88

ourke

-0.87

 accomp

-0.69

 outwe

-0.68

urance

-0.67

 fundament

-0.67

agons

-0.66

 somew

-0.66

Frameworks

-0.66

POSITIVE LOGITS

 Aden

0.82

 Lunar

0.64

 Mass

0.64

 Hemp

0.63

Nan

0.62

mone

0.62

 Cong

0.60

ICT

0.59

Nir

0.59

riot

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.