We present FinLLaMA-RAG, a Retrieval-Augmented Generation (RAG) framework built on LLaMA 3 to address complex document-level question answering. The model integrates dense retrieval, contextual fusion, and multi-hop reasoning to extract accurate, coherent answers from long and cross-referenced documents. Designed with high-stakes applications in mind, FinLLaMA-RAG supports automated tax compliance, financial fraud investigation, and regulatory reporting. Through joint optimization of retrieval and generation, it achieves strong results across several financial QA datasets. This work contributes to the development of explainable and reliable AI tools for financial and policy-oriented document analysis.

In this paper, a kernel Extreme Learning Machine (KELM) model based on vector weighted average algorithm is proposed for the prediction of national tax revenue ratio, which provides a new way of thinking and method for tax revenue prediction. By analyzing the correlation between each index and tax share, it is found that gasoline price and life expectancy are significantly positively correlated with tax share, while fertility rate and birth rate are significantly negatively correlated. The model shows excellent predictive performance on both training set and test set, with an R² of 0.995 in training set and 0.994 in test set, indicating that the model has excellent generalization ability. In addition, the root mean square error (RMSE) of the training set and the test set are 0.185 and 0.177, respectively, and the relative prediction deviation (RPD) is 14.234 and 13.178, respectively, which further verifies the high accuracy and stability of the model. Scatter plots of actual predicted versus actual values show that the model is able to accurately capture trends in tax shares with little prediction error. In summary, the optimized KELM model proposed in this paper not only has excellent performance on known data, but also has good expansion ability, and can be effectively applied to the tax share prediction of unknown data, providing a reliable tool for relevant policy making and economic analysis. The research of this paper provides a new technical path for the field of tax forecasting, which has important theoretical significance and practical value.

This paper presents a novel Retrieval-AugmentedGeneration (RAG) framework tailored for complex questionanswering tasks, addressing challenges in multi-hop reasoningand contextual understanding across lengthy documents. Builtupon LLaMA 3, the framework integrates a dense retrievalmodule with advanced context fusion and multi-hop reasoningmechanisms, enabling more accurate and coherent responsegeneration. A joint optimization strategy combining retrievallikelihood and generation cross-entropy improves the model’srobustness and adaptability. Experimental results show that theproposed system outperforms existing retrieval-augmented andgenerative baselines, confirming its effectiveness in deliveringprecise, contextually grounded answers.

In this paper, a kernel Extreme Learning Machine (KELM) model based on vector weighted average algorithm is proposed for the prediction of national tax revenue ratio, which provides a new way of thinking and method for tax revenue prediction. By analyzing the correlation between each index and tax share, it is found that gasoline price and life expectancy are significantly positively correlated with tax share, while fertility rate and birth rate are significantly negatively correlated. The model shows excellent predictive performance on both training set and test set, with an R² of 0.995 in training set and 0.994 in test set, indicating that the model has excellent generalization ability. In addition, the root mean square error (RMSE) of the training set and the test set are 0.185 and 0.177, respectively, and the relative prediction deviation (RPD) is 14.234 and 13.178, respectively, which further verifies the high accuracy and stability of the model. Scatter plots of actual predicted versus actual values show that the model is able to accurately capture trends in tax shares with little prediction error. In summary, the optimized KELM model proposed in this paper not only has excellent performance on known data, but also has good expansion ability, and can be effectively applied to the tax share prediction of unknown data, providing a reliable tool for relevant policy making and economic analysis. The research of this paper provides a new technical path for the field of tax forecasting, which has important theoretical significance and practical value.

This paper presents a novel Retrieval-Augmented Generation (RAG) framework tailored for complex question answering tasks, addressing challenges in multi-hop reasoning and contextual understanding across lengthy documents. Built upon LLaMA 3, the framework integrates a dense retrieval module with advanced context fusion and multi-hop reasoning mechanisms, enabling more accurate and coherent response generation. A joint optimization strategy combining retrieval likelihood and generation cross-entropy improves the model's robustness and adaptability. Experimental results show that the proposed system outperforms existing retrieval-augmented and generative baselines, confirming its effectiveness in delivering precise, contextually grounded answers.

This paper presents a novel Retrieval-AugmentedGeneration (RAG) framework tailored for complex questionanswering tasks, addressing challenges in multi-hop reasoningand contextual understanding across lengthy documents. Builtupon LLaMA 3, the framework integrates a dense retrievalmodule with advanced context fusion and multi-hop reasoningmechanisms, enabling more accurate and coherent responsegeneration. A joint optimization strategy combining retrievallikelihood and generation cross-entropy improves the model’srobustness and adaptability. Experimental results show that theproposed system outperforms existing retrieval-augmented andgenerative baselines, confirming its effectiveness in deliveringprecise, contextually grounded answers.