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

brush

-0.80

sonian

-0.72

Narr

-0.71

powder

-0.67

[&

-0.66

çĦ

-0.65

Quantity

-0.64

Enlarge

-0.64

Impl

-0.64

frames

-0.63

POSITIVE LOGITS

uca

0.75

vation

0.73

unia

0.66

opsy

0.65

opathic

0.62

 amalg

0.62

azon

0.61

ads

0.60

quer

0.60

icular

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.