Entity · paper

Improving LLM-Generated Process Model Quality Through Reinforcement Learning: The Role of Reward Function Design

paperactiveprovisionalimproving-llm-generated-process-model-quality-through-reinforcement-learning-the-role-of-reward-function-design-c64ee1c2·1 events·first seen Jul 8, 2026

Aliases: Improving LLM-Generated Process Model Quality Through Reinforcement Learning: The Role of Reward Function Design

Co-occurring entities

Qwen2.5-7B GSPO (Group Sequence Policy Optimization)Llama-3.1-8B

More like this (12)

Gradient-Guided Reward Optimization Using Reward Uncertainty to Induce Diverse Behaviour in Reinforcement Learning rule-based reinforcement learning rewards Learning Process Rewards via Success Visitation Matching for Efficient RL ExpRL: Exploratory RL for LLM Mid-Training Scaling Laws for Reward Model Overoptimization Reinforcement Learning with Metacognitive Feedback Reinforcement Learning with Metacognitive Feedback Elicits Faithful Uncertainty Expression in LLMs Entropy-Regularized Reinforcement Learning Visually Grounded Self-Reflection for Vision-Language Models via Reinforcement Learning Reinforcement Learning for Code Process Reward Model

Recent events (1)

5arXiv · cs.CL·Jul 8, 2026·source ↗

RL reward function design for LLM-generated BPMN process models: systematic study across 48 configurations

Researchers present a systematic study of reward function design for reinforcement learning applied to LLM-based BPMN process model generation, training Llama 3.1 8B and Qwen 2.5 14B across 48 configurations using Group Sequence Policy Optimization. Key findings: RL substantially improves syntactic and pragmatic quality while preserving semantic fidelity, equal reward weighting outperforms targeted weighting, and reward design effects interact with model architecture in non-trivial ways. The paper argues reward composition is as consequential as the decision to apply RL at all, with implications for any multi-dimensional structured generation task.

Evaluation and Benchmarking Alignment and RLHF Qwen2.5-7B Improving LLM-Generated Process Model Quality Through Reinforcement Learning: The Role of Reward Function Design GSPO (Group Sequence Policy Optimization)+1 more