🤖 AI Summary
Large language models (LLMs) exhibit fragility in complex logical reasoning due to disordered and redundant context, undermining multi-hop inference. To address this, we propose the “Concise and Ordered Perception” (COP) paradigm—the first to identify human-like cognitive biases in LLMs—and introduce three core components: semantic-driven information distillation, context-restructuring prompts, and task-adaptive reasoning chain organization. These jointly enable principled filtering and structured reorganization of critical logical information, thereby enhancing robust perception of logical dependencies. Evaluated on five rigorous logical reasoning benchmarks—ProofWriter, PrOntoQA, FOLIO, DI-GSM, and LogiQA—COP achieves average accuracy improvements of 8.2–15.6% over state-of-the-art methods. Results demonstrate COP’s effectiveness and generalizability in strengthening logical reasoning robustness across diverse formal and natural-language reasoning tasks.
📝 Abstract
Exploiting large language models (LLMs) to tackle reasoning has garnered growing attention. It still remains highly challenging to achieve satisfactory results in complex logical problems, characterized by plenty of premises within the context and requiring multi-hop reasoning. In particular, the reasoning capabilities of LLMs are brittle to disorder and distractibility. In this work, we first examine the mechanism from the perspective of information flow and reveal that LLMs confront difficulties akin to human-like cognitive biases when dealing with disordered and irrelevant content in reasoning tasks. However, in contrast to LLMs, disordered and irrelevant content does not significantly decrease human performance, as humans have a propensity to distill the most relevant information and systematically organize their thoughts, aiding them in responding to questions.Stem from that, we further propose a novel reasoning approach named Concise and Organized Perception (COP). COP carefully analyzes the given statements to identify the most pertinent information while eliminating redundancy efficiently. It then prompts the LLMs in a more organized form that adapts to the model's inference process. By perceiving concise and organized context, the reasoning abilities of LLMs can be better elicited. Extensive experimental results on several popular logical benchmarks (ProofWriter, PrOntoQA, PrOntoQA-OOD, and FOLIO) and mathematical benchmark (DI-GSM) show that COP significantly outperforms previous state-of-the-art methods.
