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

CHAT

-0.82

mop

-0.79

 Malt

-0.61

alle

-0.61

anned

-0.60

rimp

-0.57

GH

-0.57

olitan

-0.56

wcsstore

-0.56

fml

-0.56

POSITIVE LOGITS

 stereotypes

0.68

iors

0.68

 sovere

0.66

ĺħ

0.65

uably

0.65

ibel

0.64

 Publisher

0.64

 tropes

0.63

ĵĺ

0.62

 ceilings

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.