Bridging Auxiliary Constraints to Resolve Instruction Following in Large Reasoning Models
The paper introduces a framework to improve instruction following in Large Reasoning Models (LRMs) by addressing the Constraint Adherence Problem (CAP). It proposes a method called Constraint Relationship Graph Completion (CRGC) that models relationships between constraints and identifies auxiliary 'bridge constraints' to enhance instruction clarity. Experimental results show a significant reduction in constraint violations while maintaining reasoning capabilities.
- ▪Large Reasoning Models struggle with reliably following multiple instructions due to competing constraints.
- ▪The proposed framework, CRGC, models constraints as a structured knowledge graph.
- ▪Experiments indicate that CRGC reduces constraint violations by 39% compared to standard prompting.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Artificial Intelligence arXiv:2606.03624 (cs) [Submitted on 2 Jun 2026] Title:Bridging Auxiliary Constraints to Resolve Instruction Following in Large Reasoning Models Authors:Zhengyi Zhao, Shubo Zhang, Huimin Wang, Zezhong Wang, Yutian Zhao, Yefeng Zheng, Binyang Li, Yulan He, Kam-Fai Wong, Xian Wu View a PDF of the paper titled Bridging Auxiliary Constraints to Resolve Instruction Following in Large Reasoning Models, by Zhengyi Zhao and 9 other authors View PDF HTML (experimental) Abstract:Large Reasoning Models (LRMs) have demonstrated impressive capabilities in many tasks, yet they struggle with reliably following multiple instructions, either by failing to satisfy individual constraints or by struggling to balance competing constraints simultaneously.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
