INDEX

Explanations

praise or positive sentiment expressed towards oneself

oai_token-act-pair · gpt-3.5-turbo

the pronoun "I" in various contexts, indicating personal expression and sentiment

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

 provocation

-0.77

 violations

-0.76

 contamin

-0.71

 unspecified

-0.69

 jurisd

-0.69

 violating

-0.69

 delinquent

-0.68

odder

-0.66

 methodological

-0.66

 obfusc

-0.64

POSITIVE LOGITS

 congratulate

1.46

 adore

1.33

 thank

1.31

 appreciate

1.27

 love

1.22

 enjoyed

1.21

 commend

1.21

 cherish

1.20

 loved

1.19

udos

1.18

Activations Density 0.280%

praise or positive sentiment expressed towards oneself

the pronoun "I" in various contexts, indicating personal expression and sentiment

No Comments

No Known Activations