You are an expert in computer vision and explainable artificial intelligence. Your task is to analyze visual examples and discover interpretable, human-understandable features that can differentiate between two hidden image categories.
You must reason purely from visual evidence, without assuming any predefined feature types, and describe the distinguishing properties as if explaining them to a human researcher.


I am providing you with example images.
These images come from two unknown visual categories (two classes), but you are not told which image belongs to which class.

Your task is to:
	1.	Examine all images carefully and propose the key visual features that appear to systematically vary between subsets of the images — these features should be sufficient to distinguish two distinct categories if used in a classification model.
	2.	Focus on visual characteristics that are meaningful and explainable to a human observer (for example, differences in structure, function, material, context, or visual composition) — but do not predefine feature types like “color” or “shape” unless they emerge naturally from your reasoning.
	3.	Express each feature in clear, interpretable language — ideally as short phrases or attribute names (e.g., “has visible handle”, “open top”, “contains food-like surface”, etc.).
	4.	For each feature, provide a brief explanation of why it could be discriminative between two hidden groups.
	5. For each feature, provide a possible values.
	6. Provide at least 10 features.

Output format:
{
  "proposed_features": [
    {"feature": "feature_name_1", "description": "short explanation why it may separate the two hidden groups", "possible_values": ["value1", "value2", "value3", "value4", "value5"]},
    {"feature": "feature_name_2", "description": "short explanation", "possible_values": ["value1", "value2", "value3", "value4", "value5"]},
    ...
  ]
}