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

 Sutherland

-0.76

 Razor

-0.75

Hill

-0.71

 Simpson

-0.70

erella

-0.68

 Trident

-0.67

Ala

-0.67

Job

-0.67

INA

-0.65

Hom

-0.64

POSITIVE LOGITS

anwhile

0.82

ngth

0.81

 amounts

0.75

 athlet

0.73

 disposed

0.73

 inflamm

0.71

umes

0.71

anse

0.70

ontent

0.70

 neut

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.