INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Cyborg

-0.70

Book

-0.67

 Baal

-0.66

Board

-0.62

 Dawkins

-0.61

 Osiris

-0.61

Charge

-0.61

aying

-0.61

loaded

-0.61

ONEY

-0.60

POSITIVE LOGITS

soType

0.84

 recre

0.80

kef

0.78

ãĤ¦ãĤ¹

0.73

 videog

0.71

ãĥ¼ãĥĨ

0.71

 thrill

0.69

ograph

0.68

andise

0.67

quickShipAvailable

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.