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

eatures

-0.82

 psychiat

-0.76

theless

-0.75

 shenan

-0.75

 centerpiece

-0.70

awa

-0.69

 includ

-0.68

 skelet

-0.68

 clocks

-0.67

weather

-0.66

POSITIVE LOGITS

vir

0.77

 Username

0.74

BILL

0.73

DK

0.73

arian

0.72

enum

0.72

Ð´

0.70

arine

0.69

ariat

0.69

 Viktor

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.