datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
hr_assistant_qna_datasetHRE-ToM-KnowledgePerceptionThe Knowledge-Perception task aims to evaluate the ability to link perception to knowledge. Each scenario involves one person moving an object to a new location and another person leaving and later returning to the room. The second person either observes or does not observe the other person moving the object.
The following template can be used to evaluate LLMs on the Knowledge-Perception task:f"Read the scenario and answer the question by using step by step reasoning:\n\nScenario:… See the full description on the dataset page: https://huggingface.co/datasets/Z3R6X/HRE-ToM-KnowledgePerception.hr-chat-dataHRE-ToM-InferringEmotionThe Inferring Emotion task aims to evaluate the ability to infer the emotion of a person based on situational context. Each scenario describes a situation involving two persons, where one person experiences an emotion triggered by an event. This emotion can be inferred from the observable actions and the behavior of the person.
The following template can be used to evaluate LLMs on the Inferring Emotion task:f"Read the scenario and answer the question by using step by step… See the full description on the dataset page: https://huggingface.co/datasets/Z3R6X/HRE-ToM-InferringEmotion.