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

 Plat

-0.65

ĺħ

-0.64

 Baxter

-0.63

 Kenobi

-0.62

 Jets

-0.62

Vac

-0.62

 Edwards

-0.61

Hue

-0.61

 Thib

-0.60

 Cousins

-0.60

POSITIVE LOGITS

eatures

0.93

velt

0.89

ordes

0.85

izen

0.81

rieg

0.77

worm

0.73

ront

0.72

ross

0.71

ures

0.71

itizens

0.71

Activations Density 0.000%

No Known Activations

This feature has no known activations.