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

 Races

-0.75

 Proud

-0.73

 Defenders

-0.71

places

-0.70

Kry

-0.68

 religions

-0.67

llo

-0.67

 Appears

-0.66

 Hallow

-0.65

 Garrison

-0.65

POSITIVE LOGITS

-->

0.82

AMI

0.73

ibur

0.72

Motor

0.70

0.69

BIP

0.69

icz

0.68

Asset

0.67

istar

0.66

ashi

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.