The hallucination of LLM that people criticize is an artificial intelligence effort.
Haebom
Hallucination is a phenomenon in which an AI model produces incorrect or strange results when performing a specific task. This problem can occur because the model has learned to select the most likely word by considering the given words during the learning process, and because the nature of artificial neural networks (e.g. Transformer) is to compress and abstract information, and this phenomenon occurs because the information is not stored in its original form.
In Korea, this is an AI error that has become famous, such as the King Sejong MacBook throwing incident , but in fact, this is a trace of the AI trying its best due to the LLM characteristic of connecting the next word. In such cases, the question (prompt) is often incorrect or ambiguous.
It is a misnomer to call the LLM (Large Language Model) creating false information 'hallucination'. Strictly speaking, this phenomenon is more appropriately explained by the term 'confabulation', which is already used in psychology. Confabulation is a phenomenon that generally occurs when people with brain damage cannot answer questions accurately. Likewise, LLM creates plausible information by forcing it when it does not know the correct information.
How to solve hallucination in artificial intelligence
To reduce hallucination, the model is trained additionally with processed data such as domain-specific expertise and question-answer data.
We need to continuously improve our models based on user feedback. This can be supplemented by using reinforcement learning (RLHF).
Fine-tuning the model's output to generate based on the embedding AP (Answer Preserving) and the answer sheet to ensure the correct answer.
Complement the model by searching and linking with apps and plugins for specific fields, providing accurate information.
How to treat confabulation in real psychotherapy
Improving Awareness: The primary goal of Confabulation Therapy is to help patients recognize the inaccuracies of their memories. This can improve awareness and provide a better understanding of their memories and behaviors.1
Collaborative Approach: A collaborative approach can be used to provide psychological constructs to develop awareness of confabulation. This can help patients understand the causes and effects of confabulation and develop an appropriate treatment plan2.
Processing Emotions: If confabulations occur frequently, it may be important to process the emotions that cause them. This can be done through psychotherapy, and understanding and processing these emotions can reduce the frequency of confabulations.
Handling Memory Test Intrusions and Simple-Trigger Confabulations: Confabulations have been proposed to be of four types, such as intrusions in memory tests or simple-trigger confabulations , and it may be important to identify and handle these types.
Memory exercises and training: Memory exercises and training can be provided to improve memory problems related to confabulation. This can help the patient understand the difference between reality and memory, and improve the accuracy of memory.
Stress Management: Confabulation can be triggered by stress or changes in emotional state, so it may be important to learn and practice stress management techniques.
Maintain a healthy lifestyle: Maintaining a healthy lifestyle, including regular exercise, a balanced diet, and enough sleep, can improve brain health and reduce confabulation symptoms.
📚 Welcome to Haebom's archives. --- I post articles related to IT 💻, economy 💰, and humanities 🎭. If you are curious about my thoughts, perspectives or interests, please subscribe. Would you like to be notified when new articles are posted? 🔔 Yes, that means subscribe. haebom@kakao.com
