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

 Thai

-0.84

 Cambod

-0.70

unci

-0.67

itative

-0.67

ensional

-0.66

ses

-0.66

adic

-0.66

reciation

-0.65

pas

-0.63

atively

-0.62

POSITIVE LOGITS

yrinth

0.71

 millenn

0.70

 contam

0.67

 giants

0.67

 creek

0.65

 Favorite

0.62

Fac

0.62

 metab

0.61

ername

0.61

 idol

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.