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

encia

-0.81

icket

-0.78

SPONSORED

-0.76

 bats

-0.74

ank

-0.74

 destro

-0.71

 balcon

-0.68

ffield

-0.68

iband

-0.68

arth

-0.66

POSITIVE LOGITS

 plural

0.73

 Lutheran

0.72

ipeg

0.69

rite

0.68

terson

0.67

Pew

0.65

 Works

0.63

riz

0.62

ennial

0.61

 Sche

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.