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

hyde

-0.79

Ħ¢

-0.69

FactoryReloaded

-0.68

fug

-0.63

ergus

-0.62

 Tone

-0.61

\\\\

-0.59

 Haram

-0.59

etus

-0.58

utral

-0.58

POSITIVE LOGITS

ovsky

0.70

ERA

0.70

leaf

0.69

peat

0.66

norm

0.66

izont

0.62

vs

0.61

innon

0.60

 Blazers

0.58

country

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.