Simtoon ML Model Protocol v1.0

Simtoon ML Model Protocol v1.0

dated July 6, 2023

This Simtoon Machine Learning Model Protocol (the “Protocol”) establishes guidelines for the responsible utilization of machine learning models, specifically image classifiers trained on human data and language models, as defined in the OngakkenAI ML License v1.0.

Non-Discrimination: Image classifiers trained on human data must not be employed to discriminate or participate in any decision-making process resulting in differential treatment of individuals or groups.

Non-Judgement: Image classifiers trained on human data must not be used to assess individuals based solely on physical appearances. Comprehensive evaluation of individuals should consider diverse aspects, including but not limited to, personality traits, hobbies, and professional qualifications.

Integrity of Information: Language models must not be utilized with the intention to create, proliferate, or disseminate misleading or false information.

Qualified Human Supervision: Language models must not operate as standalone decision-making entities. Their operation should be under the guidance of knowledgeable human operators. Specifically, when involved in decision-making, they must be overseen by individuals sufficiently trained in the field of machine learning, with a particular emphasis on Transformer technology.

Transparency: The employment of language models must not aim to deceive or mislead intentionally. It must be evident to users when an interaction is facilitated or influenced by a language model.

Respect for Human Rights: Language models must be trained and utilized in a manner that respects fundamental human rights as defined by international law and norms. This includes, but is not limited to, the right to privacy, freedom of speech, and freedom from discrimination.

Adherence to this Protocol is required in the use, distribution, and redistribution of all machine learning models under the OngakkenAI ML License v1.0. Any deviation from this Protocol will be considered a violation of the License.