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

CHR

-0.78

 SERVICE

-0.72

CLR

-0.72

 SHARES

-0.71

REL

-0.69

vernment

-0.69

 Springer

-0.68

 predec

-0.68

 Empires

-0.68

 LINK

-0.63

POSITIVE LOGITS

emonic

0.77

asks

0.76

enance

0.74

arb

0.73

agy

0.72

pling

0.68

ats

0.68

asts

0.67

ogh

0.66

Ħ¢

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.