INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DNA

-0.68

anos

-0.65

endez

-0.65

rons

-0.64

taboola

-0.61

 glyc

-0.60

nesday

-0.60

 versus

-0.60

 differential

-0.59

ulton

-0.59

POSITIVE LOGITS

BuyableInstoreAndOnline

0.81

DragonMagazine

0.76

REDACTED

0.74

hang

0.72

atism

0.70

hare

0.70

occ

0.69

ãĥ´ãĤ¡

0.69

\",

0.67

åĤ

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.