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

 Logged

-0.69

ploma

-0.69

empl

-0.69

 Nato

-0.68

ilogy

-0.68

Pip

-0.68

 endeavour

-0.67

 civilisation

-0.64

Ïī

-0.64

 endeav

-0.64

POSITIVE LOGITS

 Marqu

0.77

ARI

0.75

haw

0.72

Der

0.71

 Barron

0.70

Hoy

0.70

Dia

0.66

eat

0.65

Zur

0.65

 Debor

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.