Language models improve math and logic tasks with creative reasoning
MIRAGE: Multi-Perspective Creative Language Model Reasoning with Reinforcement Learning Guidance
Computation and Language
Summary
Large language models have trouble solving complicated math and logical problems. The authors designed MIRAGE, a method that helps the model think in different ways, like switching strategies to solve a puzzle. MIRAGE picks good ways to approach a problem and tries solving it step-by-step, combining multiple ideas if needed. This approach leads to better accuracy on math and reasoning tests without much extra computing time.
What this means in practice
- •For software developers: Build AI assistants that solve complex math and logic problems more accurately in user applications.
- •For financial analysts: Improve automatic reasoning tools for probabilistic and algebraic problem solving in financial modeling.
Authors
Arash Lagzian, Srinivas Anumasa, Dianbo Liu
Abstract
Recent advances in Large Language Models (LLMs) have revolutionized artificial intelligence and how human interact with AIs. Despite impressive advancements, LLMs struggle with complex mathematical, scientific, and logical tasks. Inspired by human cognitive flexibility - our ability to dynamically switch mental perspectives - we propose MIRAGE (Multi-perspective Inference-time Reasoning via Agent-Guided Exploration), a novel inference-time creative thinking framework. MIRAGE includes a Selector that prioritizes effective conceptual perspectives (e.g., algebraic, probabilistic) and a Reasoner that sequentially solves tasks until a confident solution emerges, otherwise aggregating multiple perspectives. Tested on GSM8K, MATH500, MMLU-Pro, and Game-of-24 benchmarks, MIRAGE consistently outperforms methods like Chain-of-Thought and diverse prompting ensembles, significantly boosting accuracy with minimal inference overhead, providing a scalable solution for practical applications.