This paper presents a comprehensive survey of over thirty-two techniques developed to mitigate hallucination in large language models (LLMs). Hallucination refers to the generation of factually incorrect or ungrounded information by LLMs, which poses a significant challenge for their safe deployment in real-world applications. The paper introduces a detailed taxonomy of hallucination mitigation techniques, categorizing them based on various parameters such as dataset utilization, common tasks, feedback mechanisms, and retriever types. This classification helps distinguish the diverse approaches specifically designed to tackle hallucination issues in LLMs. The paper also analyzes the challenges and limitations inherent in these techniques, providing a solid foundation for future research in addressing hallucinations and related phenomena within the realm of LLMs. The survey covers various techniques for hallucination mitigation, including retrieval-augmented generation (RAG), knowledge retrieval, and self-refinement through feedback and reasoning. RAG enhances the responses of LLMs by tapping into external, authoritative knowledge bases rather than relying on potentially outdated training data or the model's internal knowledge. This approach addresses the key challenges of accuracy and currency in LLM outputs. RAG effectively mitigates the issue of hallucination in LLMs by generating responses that are not only pertinent and current but also verifiable, thereby reinforcing user confidence and offering developers an economical way to enhance the fidelity and utility of LLMs across different applications. Prompt engineering is another important technique for hallucination mitigation, involving the use of specific instructions to guide the model's output. Techniques such as retrieval augmented generation, knowledge retrieval, and real-time verification and rectification are discussed in detail. These methods aim to reduce the occurrence of hallucinations by incorporating external knowledge, validating generated content, and ensuring the accuracy of the model's outputs. The paper also explores the development of novel models to mitigate hallucinations, including the use of knowledge graphs and the introduction of faithfulness-based loss functions. These approaches aim to enhance the model's ability to generate accurate and reliable information by incorporating external knowledge and refining the model's training process. Overall, the paper provides a comprehensive overview of the various techniques used to mitigate hallucination in LLMs, highlighting their effectiveness, limitations, and potential for future research. The survey underscores the importance of addressing hallucination in LLMs due to their integral role in critical tasks and the need for robust and reliable language generation systems.This paper presents a comprehensive survey of over thirty-two techniques developed to mitigate hallucination in large language models (LLMs). Hallucination refers to the generation of factually incorrect or ungrounded information by LLMs, which poses a significant challenge for their safe deployment in real-world applications. The paper introduces a detailed taxonomy of hallucination mitigation techniques, categorizing them based on various parameters such as dataset utilization, common tasks, feedback mechanisms, and retriever types. This classification helps distinguish the diverse approaches specifically designed to tackle hallucination issues in LLMs. The paper also analyzes the challenges and limitations inherent in these techniques, providing a solid foundation for future research in addressing hallucinations and related phenomena within the realm of LLMs. The survey covers various techniques for hallucination mitigation, including retrieval-augmented generation (RAG), knowledge retrieval, and self-refinement through feedback and reasoning. RAG enhances the responses of LLMs by tapping into external, authoritative knowledge bases rather than relying on potentially outdated training data or the model's internal knowledge. This approach addresses the key challenges of accuracy and currency in LLM outputs. RAG effectively mitigates the issue of hallucination in LLMs by generating responses that are not only pertinent and current but also verifiable, thereby reinforcing user confidence and offering developers an economical way to enhance the fidelity and utility of LLMs across different applications. Prompt engineering is another important technique for hallucination mitigation, involving the use of specific instructions to guide the model's output. Techniques such as retrieval augmented generation, knowledge retrieval, and real-time verification and rectification are discussed in detail. These methods aim to reduce the occurrence of hallucinations by incorporating external knowledge, validating generated content, and ensuring the accuracy of the model's outputs. The paper also explores the development of novel models to mitigate hallucinations, including the use of knowledge graphs and the introduction of faithfulness-based loss functions. These approaches aim to enhance the model's ability to generate accurate and reliable information by incorporating external knowledge and refining the model's training process. Overall, the paper provides a comprehensive overview of the various techniques used to mitigate hallucination in LLMs, highlighting their effectiveness, limitations, and potential for future research. The survey underscores the importance of addressing hallucination in LLMs due to their integral role in critical tasks and the need for robust and reliable language generation systems.