ChatDB: Augmenting LLMs with Databases as Their Symbolic Memory
This addresses the issue of error accumulation in neural memory for LLMs, enabling more reliable complex reasoning, though it is incremental as it builds on existing computer architecture concepts.
The authors tackled the problem of LLMs' limited memory for complex reasoning by augmenting them with symbolic memory using SQL databases, resulting in a framework where the LLM generates SQL instructions to manipulate databases, validated on a synthetic dataset requiring complex multi-hop reasoning.
Large language models (LLMs) with memory are computationally universal. However, mainstream LLMs are not taking full advantage of memory, and the designs are heavily influenced by biological brains. Due to their approximate nature and proneness to the accumulation of errors, conventional neural memory mechanisms cannot support LLMs to simulate complex reasoning. In this paper, we seek inspiration from modern computer architectures to augment LLMs with symbolic memory for complex multi-hop reasoning. Such a symbolic memory framework is instantiated as an LLM and a set of SQL databases, where the LLM generates SQL instructions to manipulate the SQL databases. We validate the effectiveness of the proposed memory framework on a synthetic dataset requiring complex reasoning. The project website is available at https://chatdatabase.github.io/ .