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

hyd

-0.83

essential

-0.75

wage

-0.65

 worker

-0.65

vernight

-0.65

emia

-0.64

umph

-0.64

azar

-0.64

 cyan

-0.62

 cycl

-0.62

POSITIVE LOGITS

 Remastered

0.87

 Presents

0.76

 Forsaken

0.75

 mathemat

0.71

Ct

0.67

 Rivals

0.66

 Drawn

0.65

 EDITION

0.64

 GOODMAN

0.64

 Guides

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.