INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Pengu

-0.69

 Elev

-0.68

 elev

-0.65

 pneum

-0.62

lessness

-0.61

ailability

-0.61

osity

-0.61

 Intake

-0.61

 Temper

-0.60

 cement

-0.60

POSITIVE LOGITS

natureconservancy

0.79

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

0.68

audi

0.68

rique

0.67

âĹ¼

0.64

eq

0.64

 Appears

0.61

arist

0.61

bench

0.60

forces

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.