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

rament

-0.85

=-=-=-=-=-=-=-=-

-0.81

ULTS

-0.76

ftime

-0.76

 halftime

-0.72

quickShipAvailable

-0.71

liga

-0.69

 shooters

-0.68

eland

-0.67

 Dunham

-0.66

POSITIVE LOGITS

 Dise

0.68

 unsub

0.66

uyomi

0.63

arat

0.61

uc

0.60

 Antar

0.59

oso

0.58

aro

0.58

aber

0.58

farious

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.