Mustafa Anis Hussain
1 indexed paper
Recent (6 mo)
1With code
0Influential cites
0Benchmarked
0Research Timeline
2026
Planner-Centric Reinforcement Learning for Deep Research with Structure-Aware Reward
The paper proposes DecomposeR, a planner-centric framework that structures deep research into typed Directed Acyclic Graphs (DAGs) to explicitly improve the planning and execution of large language models for complex, multi-branch inquiries.
Highlighted terms show continued research focus across papers