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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pillar

-0.80

Downloadha

-0.78

isSpecialOrderable

-0.77

 tiss

-0.69

soDeliveryDate

-0.68

apult

-0.68

Fed

-0.65

odynam

-0.65

 scen

-0.64

 territ

-0.63

POSITIVE LOGITS

Boo

0.68

Write

0.67

 millionaires

0.63

 Responsibility

0.62

enna

0.62

elle

0.61

aru

0.60

Boo

0.60

ude

0.60

emp

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.