Reducing Reasoning Costs: The Path of Optimization for Chain of Thought via Sparse Attention Mechanism
This work addresses computational efficiency for users of large language models, though it appears incremental as it applies a known technique (sparse attention) to a specific bottleneck (chain-of-thought reasoning).
The research tackled the high inference costs of chain-of-thought reasoning in large language models by proposing a sparse attention mechanism that focuses on few relevant tokens, resulting in significantly lower reasoning time and chain-of-thought length compared to o1 Preview on MIT linear algebra tests.
In order to address the chain of thought in the large language model inference cost surge, this research proposes to use a sparse attention mechanism that only focuses on a few relevant tokens. The researcher constructed a new attention mechanism and used GiantRabbit trained with custom GPTs as an experimental tool. The experiment tested and compared the reasoning time, correctness score and chain of thought length of this model and o1 Preview in solving the linear algebra test questions of MIT OpenCourseWare. The results show that GiantRabbit's reasoning time and chain of thought length are significantly lower than o1 Preview. It verifies the feasibility of sparse attention mechanism for optimizing chain of thought reasoning. Detailed architectural details and experimental process have been uploaded to Github, the link is:https://github.com/brucewang123456789/GeniusTrail.git.