INDEX

Explanations

verbs indicating future or potential actions

oai_token-act-pair · gpt-3.5-turbo

modal verbs indicating possibility or necessity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

stood

-0.67

 Outside

-0.60

Wr

-0.58

 Maker

-0.58

 Olymp

-0.56

 Azerb

-0.55

 Habit

-0.54

edIn

-0.53

esides

-0.53

 First

-0.53

POSITIVE LOGITS

ync

0.93

ĸļ

0.87

yip

0.81

 imply

0.78

atem

0.78

"]=>

0.71

 overwrite

0.71

 enable

0.71

ratom

0.71

 rejoice

0.71

Activations Density 0.076%

verbs indicating future or potential actions

modal verbs indicating possibility or necessity

No Comments

No Known Activations