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

 Wolverine

-0.68

)]

-0.68

 Vera

-0.68

 Sawyer

-0.65

 Ezek

-0.62

 Geral

-0.61

 Sisters

-0.61

bip

-0.60

 Mothers

-0.60

 Weston

-0.59

POSITIVE LOGITS

ifice

0.83

corruption

0.74

iculty

0.70

nown

0.70

rito

0.69

cffff

0.69

à¼

0.69

haust

0.68

ustom

0.68

fortable

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.