INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

lace

-0.73

cham

-0.68

aug

-0.67

sor

-0.67

eties

-0.66

Justice

-0.64

 glyph

-0.64

 fixme

-0.61

Round

-0.61

 GEAR

-0.61

POSITIVE LOGITS

 ende

0.71

 neighbours

0.68

 divides

0.66

creen

0.62

 eBook

0.62

 Rebell

0.61

 compuls

0.61

leneck

0.61

 Tanz

0.61

zee

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.