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

bang

-0.78

ilan

-0.68

ashtra

-0.68

ModLoader

-0.67

guns

-0.64

actory

-0.64

eners

-0.64

ratom

-0.64

eneg

-0.63

gru

-0.63

POSITIVE LOGITS

 span

0.72

enture

0.70

taboola

0.69

eele

0.66

 Bret

0.65

LEASE

0.63

Der

0.63

 Mare

0.61

 yours

0.61

hov

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.