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

 stride

-0.70

 sheds

-0.68

izons

-0.67

eatures

-0.66

needs

-0.66

 awoken

-0.65

 armour

-0.64

iets

-0.64

 sinks

-0.63

 responds

-0.62

POSITIVE LOGITS

hur

0.80

tu

0.76

Jam

0.73

hall

0.70

omination

0.69

het

0.66

 import

0.65

hee

0.64

steen

0.64

ule

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.