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

glomer

-0.78

entric

-0.76

aminer

-0.75

anan

-0.73

urrent

-0.70

ille

-0.68

mares

-0.67

culosis

-0.65

 territ

-0.65

rament

-0.65

POSITIVE LOGITS

 heavy

0.71

WER

0.70

ZA

0.65

Sto

0.63

esson

0.62

OIL

0.62

KI

0.62

 Refresh

0.61

rounder

0.61

 Avenger

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.