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

ASS

-0.72

rien

-0.70

rations

-0.64

elect

-0.64

rig

-0.62

AT

-0.61

 Idea

-0.60

 ideal

-0.60

 kidding

-0.60

 attain

-0.60

POSITIVE LOGITS

cum

0.71

 Pharaoh

0.70

Ã±o

0.67

ucer

0.64

 Boots

0.64

gio

0.62

estro

0.62

 handcuffs

0.62

 Kuwait

0.62

Loader

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.