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

ipal

-0.80

uilding

-0.76

ihilation

-0.73

 conduc

-0.73

 pestic

-0.72

gpu

-0.72

cling

-0.71

abal

-0.71

 mathemat

-0.70

Oracle

-0.68

POSITIVE LOGITS

 Sessions

0.67

 Member

0.65

SAM

0.62

 Telegram

0.62

 prank

0.62

 Mash

0.62

ms

0.61

Twe

0.61

 Ticket

0.60

 whom

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.