INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 CLSID

-0.80

*/

-0.70

 sqor

-0.68

KS

-0.67

utenberg

-0.66

 Extensions

-0.66

catentry

-0.66

 Coverage

-0.64

SHIP

-0.64

trop

-0.63

POSITIVE LOGITS

lain

0.79

 prized

0.69

 stray

0.66

 discarded

0.66

 aisle

0.65

ilion

0.64

Ļ½

0.64

 scalp

0.63

illance

0.63

onite

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.