INDEX

Explanations

references to the word "you" and its related forms, indicating a focus on direct address or interaction

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 betweenstory

-0.86

styleType

-0.78

IsMutable

-0.76

ⓧ

-0.72

 ddelweddau

-0.71

uxxxx

-0.64

 يتيمه

-0.64

ConstraintMaker

-0.64

WriteBarrier

-0.63

()?;

-0.63

POSITIVE LOGITS

you

0.62

YOU

0.57

you

0.55

 yourselves

0.54

addContainerGap

0.52

 Aérea

0.52

اء

0.51

crapers

0.50

 flattery

0.49

 yours

0.48

Activations Density 0.202%

references to the word "you" and its related forms, indicating a focus on direct address or interaction

No Comments

No Known Activations