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

aceae

-0.73

rists

-0.69

 amounted

-0.68

ounge

-0.66

ique

-0.65

ishy

-0.64

abuse

-0.63

iques

-0.63

nikov

-0.61

otta

-0.61

POSITIVE LOGITS

 Mile

0.69

 Dome

0.67

ģĸ

0.67

TPPStreamerBot

0.66

 Falcons

0.66

ple

0.66

par

0.65

 Patriot

0.65

 Faces

0.65

 Millennium

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.