12 ChatGPT Secret Codes to access the hidden features

The secret codes might behave little differently depending on which version of ChatGPT you are using . It will also work with chatbots using GPT API. Let me know if you try another chatbot besides ChatGPT. There might be some that don’t work. But it would be fun to try them out.

Temperature (0 to 1):

Controls the randomness and creativity of the generated text. Lower values result in more focused outputs, while higher values lead to more varied and creative text.

Example:

temp: 0.7 
Tell me a story about a magical kingdom.

Diversity_penalty (0 to 2):

Penalizes tokens that have already been used in the response, ensuring more diverse text. Higher values increase diversity, while lower values may result in more repetitive phrases .

Example:

div_penalty: 1.5 
Write a paragraph with diverse vocabulary.

Max tokens (positive integer):

Sets the maximum length of the generated text, controlling the output size. Set a limit based on the desired word or character count.

Example:

max_tokens: 50 
Describe a beautiful sunset.

Top_p (0–1):

Sets the probability mass cutoff for token sampling, affecting the breadth of options the AI considers. Higher values introduce more randomness, while lower values lead to more focused outputs.

Example:

top_p: 0.8 
Write a creative description of a forest.

Frequency_penalty (-2 to 2):

Penalizes frequently occurring tokens in the generated text, encouraging the AI to use less common words. Positive values promote rarer words, while negative values allow for more common language.

Example:

freq_penalty: 1 
Write a sentence using uncommon words.

Presence_penalty (-2 to 2):

Penalizes tokens that have already been used in the input. This affects how often the AI repeats phrases from the input. Positive values discourage repetition, while negative values may result in echoing input text .

Example:

pres_penalty: -1 
Echo the following sentence: The quick brown fox jumps over the lazy dog.

Best_of (positive integer):

Generates a specified number of responses and returns the one with the highest score based on the model’s internal evaluation Useful for selecting a high-quality output among multiple alternatives.

Example:

best_of: 3
What are the benefits of exercise?

N (positive integer):

Determines the number of alternative completions to generate for a given input. Can be used to explore multiple ideas or variations for a prompt.

Example:

n: 4
Suggest some creative ways to use ChatGPT.

Echo (boolean: true or false):

Enables or disables the AI’s ability to echo parts of the input in the response. When set to true, the model is more likely to include phrases from the input in the generated text.

Example:

echo: true
Tell me about the importance of a balanced diet.

Stop_sequences (list of strings):

Specifies a list of strings that, when the AI encounters them, will stop the text generation. Useful for controlling the content or structure of the generated text.

Example:

stop_sequences: ["Conclusion:", "In summary,"]
Explain the process of photosynthesis.

User_attributes (list of key-value pairs):

Allows you to provide additional information about the user . Can be used by the AI to tailor the generated text according to user preferences, demographics, or other characteristics.

Example:

user_attributes: {"age": "30", "interests": ["technology", "travel"]}
What are some popular vacation destinations for tech enthusiasts?

System_level_settings (dictionary):

Gives advanced options for fine-tuning the AI’s behavior . Can be used to change the model’s internal memory, processing speed, or resource use.

Example:

system_level_settings: {"memory": "high", "processing_speed": "fast"}
Analyze the impact of artificial intelligence on the job market.