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

bourg

-0.73

 Masquerade

-0.69

 taps

-0.68

pheus

-0.67

Enlarge

-0.66

Zig

-0.66

phrine

-0.65

Pg

-0.64

 Xiang

-0.64

LV

-0.64

POSITIVE LOGITS

ensed

0.75

inately

0.73

ording

0.69

ogn

0.64

alysed

0.63

eal

0.63

urance

0.63

 conglomer

0.62

abulary

0.62

Ĥİ

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.