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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

utation

-0.70

unes

-0.63

binding

-0.63

species

-0.62

Pixel

-0.62

Bio

-0.62

perm

-0.61

utations

-0.61

adr

-0.61

une

-0.61

POSITIVE LOGITS

yip

0.87

hower

0.78

 Azerb

0.74

 exha

0.74

 Fulton

0.70

 Hust

0.69

 Curtis

0.65

 Stacy

0.64

hetto

0.64

 Gott

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.