Search papers, labs, and topics across Lattice.
This paper introduces TopoAgent, a self-evolving topological framework designed to enhance multimodal scientific reasoning by addressing the limitations of traditional linear planning. By employing a front-end decomposer to break down complex queries into visually-grounded atoms organized in a Directed Acyclic Graph (DAG), TopoAgent ensures strict context isolation and mitigates issues like visual-semantic misalignment and long-context hallucinations. Experimental results across various scientific domains reveal that TopoAgent outperforms existing linear agent frameworks, demonstrating its robustness and adaptability in autonomous reasoning tasks.
TopoAgent's innovative use of dynamic graph evolution allows for noise-resistant scientific reasoning that outperforms traditional linear models.
While Multimodal Large Language Models (MLLMs) excel in general tasks, rigorous scientific reasoning remains challenging due to the limitations of monolithic, linear planning. Such sequential designs often suffer from visual-semantic misalignment, long-context hallucinations, and brittle execution under fixed task granularity. We propose TopoAgent, a self-evolving topological framework that replaces linear trajectories with dynamic, state-isolated graph evolution. TopoAgent first employs a front-end decomposer to fracture complex queries into visually-grounded atoms. These atoms are organized into a Directed Acyclic Graph (DAG) based on their dependencies, enabling strict context isolation to shield the reasoning engine from irrelevant historical noise. Furthermore, we introduce adaptive atomic fission, which dynamically splits bottleneck nodes into finer-grained sub-atoms at runtime when tool capability boundaries are exceeded. Extensive experiments across mathematics, physics, and chemistry benchmarks demonstrate that TopoAgent significantly outperforms state-of-the-art linear agent frameworks, providing a robust, noise-resistant, and self-correcting paradigm for autonomous scientific reasoning.
