{"slug": "a-tricky-method-for-nlp-and-llm-from-my-previous-research-work", "title": "A tricky method for NLP and LLM from my previous research work", "summary": "A researcher proposed a method for NLP and LLMs that uses reverse training and inference to detect logical errors in text, such as university student reports. The approach compares normal and reverse paths to identify logical skips or errors, potentially reducing hallucination in LLMs. The method is freely available for use.", "body_md": "This post is from my research work in NLP.\n\n**-The background of this work.**\n\nI did a research on an AI system that can help professors to test reports of university students.\n\nThe main task of this system was to find out whether a report is correct and well organized in logic.\n\nThis system was supposed to give outputs such as: \"There is a logical skip between paragraph C and D \", “There is a logical error from Para B to Para E.”…\n\nMy idea was this:\n\n**-Logical base**\n\nThe mathematical(or logical) analysis or model of this problem is about Necessary condition and Sufficient condition. (I am pretty sure that every friend in this community is aware of this logic.) Here comes my main idea or trick.\n\nI took the normal training process and inferencing in models as A->B(Necessary condition). Then what about B->A(Sufficient condition)? The answer: Reverse(In my team we called this reverse-from-result training or simply reverse-result-training) training and inferencing was taken as B->A(Sufficient condition).\n\nThis means that training or inferencing first follow the normal path then continues from the normal path result to the beginning of normal path. Then the model can compare reports(inputs) and inferences. For example the similarity is bigger than 0.8, the report is logically fine. If similarity of paragraph C of the report is 0.7-0.5, there is a logical skip after paragraph C of the report. If similarity is less than 0.5, there is a logical error from Para C to D of the report.\n\n**-Conclusion**\n\nThis method can be developed into a new mechanism or replace a part of LLM model to reduce Hallucination in LLMs.\n\nFeel free to use this idea and method for any purpose since there will be no copyrights conflicts for this idea and I am open to any discussions and feedbacks on this method.", "url": "https://wpnews.pro/news/a-tricky-method-for-nlp-and-llm-from-my-previous-research-work", "canonical_source": "https://discuss.huggingface.co/t/a-tricky-method-for-nlp-and-llm-from-my-previous-research-work/177579#post_1", "published_at": "2026-07-08 13:35:45+00:00", "updated_at": "2026-07-08 13:40:44.768752+00:00", "lang": "en", "topics": ["natural-language-processing", "large-language-models", "ai-research"], "entities": [], "alternates": {"html": "https://wpnews.pro/news/a-tricky-method-for-nlp-and-llm-from-my-previous-research-work", "markdown": "https://wpnews.pro/news/a-tricky-method-for-nlp-and-llm-from-my-previous-research-work.md", "text": "https://wpnews.pro/news/a-tricky-method-for-nlp-and-llm-from-my-previous-research-work.txt", "jsonld": "https://wpnews.pro/news/a-tricky-method-for-nlp-and-llm-from-my-previous-research-work.jsonld"}}