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

 Bridgewater

-0.68

 Marble

-0.66

properties

-0.64

 Carth

-0.62

 Rivers

-0.62

 Moor

-0.61

 Leilan

-0.61

 escapes

-0.61

izz

-0.59

 Byzantine

-0.59

POSITIVE LOGITS

edIn

0.78

ega

0.73

lights

0.69

nic

0.68

Offline

0.67

nation

0.66

 Emblem

0.65

Dom

0.65

sin

0.64

Company

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.