INDEX

Explanations

expressions related to values and principles, especially in context to integrity and community service

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 35-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_35/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.35.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

الإنجليزية

-0.43

 interlocutor

-0.41

 раск

-0.41

Carriera

-0.40

 חיצוניים

-0.39

 überrascht

-0.39

 erleich

-0.39

Odds

-0.38

imli

-0.38

汰

-0.38

POSITIVE LOGITS

 principles

1.91

principles

1.63

 principle

1.61

 values

1.59

 Principles

1.52

Principles

1.49

 PRINCIPLES

1.43

 princípios

1.41

principle

1.39

 ideals

1.38

Activations Density 0.544%

expressions related to values and principles, especially in context to integrity and community service

No Comments

No Known Activations