INDEX

Explanations

references to self and user identity in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 chief

-0.36

 Chief

-0.34

 tuong

-0.34

 efectivo

-0.32

 CHIEF

-0.31

 THINGS

-0.31

 Letter

-0.30

 thẻ

-0.29

 zare

-0.29

 Effective

-0.28

POSITIVE LOGITS

own

0.69

DIY

0.68

+#+

0.65

 iprot

0.63

DIY

0.62

 thyself

0.61

diy

0.60

自分で

0.60

odziel

0.60

 włas

0.59

Activations Density 0.713%

references to self and user identity in various contexts

No Comments

No Known Activations