analiza aiqbench
This paper tests a parameter-efficient RAG pipeline built on a 4-bit-quantized 14B DeepSeek-R1 model, with layout-aware document extraction, for Indian Chartered Accountancy questions. Its reported result is about 68.75% of GPT-4o and Claude 3.5 Sonnet on the authors' own CA-Ben benchmark, which is a useful data point for cost-constrained deployments but not parity with frontier models. The authors' admission that the system struggles with complex regulatory text such as taxation is the key caveat for anyone considering domain-specific finance applications.
AI-Enabled Question-Answering for Chartered Accountancy: A Retrieval-Augmented Reasoning Framework
arXiv:2605.00257v2 Announce Type: replace-cross Abstract: The inception of Large Language Models (LLMs) has catalyzed AI adoption in the finance sector, yet their reliability in complex, jurisdiction-specific tasks like Indian Chartered Accountancy (CA) remains limited. The models display difficulty in executing numerical tasks which require multiple steps while also needing advanced knowledge about legal regulations and the method of scaling their operations is not feasible in settings which have limited access to resources. We present CA-ThinkFlow as a parameter-efficient Retrieval-Augmented Generation (RAG) framework which operates with a 14B, 4-bit-quantized reasoning model, 14B-DeepSeek-R1, and a layout-aware Docling extraction system which maintains document structure during extraction. CA-ThinkFlow uses a basic RAG method which automatically adds retrieved information into the prompt, while it depends on the model's built-in Chain-of-Thought (CoT) functions to create context and produce correct answers. The system we developed system operates at performance levels which match large proprietary models when we tested it on the multi-level CA-Ben benchmark, achieving Scholastic Reliability Coefficient (SRC) results which equal 68.75\% of GPT-4o and Claude 3.5 Sonnet. The framework shows high efficiency and strength in handling parameters, but essential reasoning abilities fail to process complex regulatory texts which exist in fields such as Taxation.
Citește articolul complet pe arXiv