Co-Sight: Enhancing LLM-Based Agents via Conflict-Aware Meta-Verification and Trustworthy Reasoning with Structured Facts

TL;DR

Co-Sight introduces a conflict-aware verification framework with structured facts to improve the reliability and transparency of long-horizon reasoning in LLM-based agents, achieving state-of-the-art accuracy and efficiency.

Contribution

It presents a novel conflict-aware meta-verification method combined with structured factual reasoning to enhance trustworthiness and efficiency in LLM reasoning processes.

Findings

Achieves 84.4% accuracy on GAIA dataset.

Attains 93.8% accuracy on Chinese-SimpleQA.

Demonstrates improved efficiency and reliability through conflict-focused verification.

Abstract

Long-horizon reasoning in LLM-based agents often fails not from generative weakness but from insufficient verification of intermediate reasoning. Co-Sight addresses this challenge by turning reasoning into a falsifiable and auditable process through two complementary mechanisms: Conflict-Aware Meta-Verification (CAMV) and Trustworthy Reasoning with Structured Facts (TRSF). CAMV reformulates verification as conflict identification and targeted falsification, allocating computation only to disagreement hotspots among expert agents rather than to full reasoning chains. This bounds verification cost to the number of inconsistencies and improves efficiency and reliability. TRSF continuously organizes, validates, and synchronizes evidence across agents through a structured facts module. By maintaining verified, traceable, and auditable knowledge, it ensures that all reasoning is grounded in consistent, source-verified information and supports transparent verification throughout the reasoning process. Together, TRSF and CAMV form a closed verification loop, where TRSF supplies structured facts and CAMV selectively falsifies or reinforces them, yielding transparent and trustworthy reasoning. Empirically, Co-Sight achieves state-of-the-art accuracy on GAIA (84.4%) and Humanity's Last Exam (35.5%), and strong results on Chinese-SimpleQA (93.8%). Ablation studies confirm that the synergy between structured factual grounding and conflict-aware verification drives these improvements. Co-Sight thus offers a scalable paradigm for reliable long-horizon reasoning in LLM-based agents. Code is available at https://github.com/ZTE-AICloud/Co-Sight/tree/cosight2.0_benchmarks.