INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

cms

-0.84

 Wanted

-0.71

SourceFile

-0.69

cape

-0.69

DOI

-0.68

 Browse

-0.68

adelphia

-0.67

pora

-0.67

STAR

-0.66

Wak

-0.64

POSITIVE LOGITS

 destro

0.77

 traged

0.73

 nostalg

0.70

 shorth

0.68

 psychiat

0.66

 detachment

0.65

 indisp

0.64

agn

0.63

 withd

0.63

 Sagan

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.