INDEX

Explanations

phrases indicating additional information or elements being included

oai_token-act-pair · gpt-3.5-turbo

instances of the word "addition" related to various contexts or topics

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

cha

-0.68

ards

-0.66

SPONSORED

-0.64

abe

-0.64

cies

-0.61

eters

-0.60

alled

-0.59

chu

-0.59

 gripped

-0.58

 peacefully

-0.58

POSITIVE LOGITS

 addition

3.51

 additions

2.25

 inclusion

1.49

 lieu

1.34

 introduction

1.28

 omission

1.22

 added

1.21

 arrival

1.19

 combination

1.17

 conjunction

1.14

Activations Density 0.021%

phrases indicating additional information or elements being included

instances of the word "addition" related to various contexts or topics

No Comments

No Known Activations