INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 insign

-0.66

Alias

-0.66

 metic

-0.63

sum

-0.63

»Ĵ

-0.61

¬¼

-0.61

 chants

-0.60

 greeting

-0.60

 newcomer

-0.58

cffffcc

-0.58

POSITIVE LOGITS

eds

0.76

ibraries

0.73

Topics

0.71

plet

0.70

ogie

0.69

Delta

0.68

Pokemon

0.68

independent

0.68

Adams

0.67

okemon

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.