Case sensitivity in mistralai/Mistral-7B-v0.1?

Beginners
1

I am seeing very differnt output just from the capitalization of “France” in “What is the capital of France?”. Why would this be the case?
PS. The code below gets formatted in a manner that messes up the space. I do not know how to fix that.
Sorry ;-(

import requests
hf_token = “YOUR_TOKEN”
completion_url2 = “https://api-inference.huggingface.co/models/”
completion_url2 += “mistralai/Mistral-7B-v0.1”
def query2(text):
response = requests.post(
completion_url2,
headers={“Authorization”: f"Bearer {hf_token}“},
json={“inputs”: text})
if response.status_code != 200:
raise ValueError(f"Request failed with status code {response.status_code}: {response.text}”)
json = response.json()
print(json)
return json

‘’’ Comment out one of these lines
query2(“What is the capital of France”);
query2(“What is the capital of france”);

‘’’
[{‘generated_text’: ‘What is the capital of France?\n\nParis is the capital of France.\n\nWhat is the capital of the United States?\n\nWashington, D.C. is the capital of the United States.\n\nWhat is the capital of the United Kingdom?\n\nLondon is the capital of the United Kingdom.\n\nWhat is the capital of Canada?\n\nOttawa is the capital of Canada.\n\nWhat is the capital of Australia?\n\nCanberra is the capital of Australia’}]

[{‘generated_text’: ‘What is the capital of france?\n\nParis is the capital of France. It is located in the north of the country, on the river Seine. It is the largest city in France, with a population of over 2 million people. Paris is known for its beautiful architecture, including the Eiffel Tower, the Louvre Museum, and the Arc de Triomphe. It is also a major center for fashion, art, and culture.\n\nWhat is the capital of france in french?\n’}]
‘’’

2

Case sensitivity plays a role, in language models. These models are trained on datasets where the capitalization of words carries meanings and contexts. For example when “France” is capitalized it is commonly understood as referring to the country itself. On the hand when it is written in lowercase as "france " it may not immediately be recognized as a noun and can be interpreted differently.

The training data of language models encompasses instances where proper nouns are capitalized. As a result the model learns to associate capitalized words like “France” with nouns specifically referring to the country. When encountering “france” written in lowercase the model might interpret it as a usage and generate a different response pattern.

To generate responses, complex algorithms and heuristics are employed by models such, as Mistral 7B v0.1. These heuristics consider capitalized words differently which can influence the output of the model.

Queries written in lowercase like “france” introduce ambiguity. In cases the model may spend time providing context regarding location or significance to clarify what is being referred to before directly answering the query assuming that users already recognize “France” as a country.