INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 genders

-0.70

misc

-0.70

ARCH

-0.67

 ILCS

-0.65

arl

-0.61

usc

-0.60

 sizes

-0.58

bay

-0.58

 Arch

-0.58

 tolerate

-0.58

POSITIVE LOGITS

aturday

0.71

maker

0.68

esson

0.65

iod

0.64

 Nolan

0.64

theless

0.64

0.63

ertodd

0.63

 Blockchain

0.63

stack

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.