5arXiv cs.CL (Computation and Language)·
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.
Related guides (2)
Related events (8)
Progress Advantage: Annotation-Free Step-Level Scoring for LLM Agents via RL Post-Training
Researchers introduce 'progress advantage,' a method that derives implicit step-level reward signals for LLM agents directly from the log-probability ratio between an RL-trained policy and its reference policy, without requiring dedicated process reward model training. The approach is shown to recover the optimal advantage function under a general stochastic MDP formulation, making it annotation-free and domain-agnostic. Validated across five benchmarks and four model families on tasks including test-time scaling, uncertainty quantification, and failure attribution, it outperforms confidence-based baselines and even dedicated trained reward models. The result is practically significant because building process reward models for agentic settings is currently a major bottleneck.
Scaling Laws for Reward Model Overoptimization
OpenAI published research investigating how reward model overoptimization scales with policy and reward model size in RLHF pipelines. The work characterizes the relationship between KL divergence from the initial policy and gold-standard reward, finding predictable degradation patterns as optimization pressure increases. This provides empirical grounding for understanding Goodhart's Law dynamics in language model fine-tuning and has implications for designing safer, more robust RLHF training regimes.
ExpRL: RL-based mid-training using human QA data as reward scaffolds for LLM reasoning
ExpRL proposes an automated approach to LLM mid-training that replaces manually curated reasoning traces with large corpora of human-written QA data used as reward scaffolds rather than imitation targets. Reference solutions are hidden from the policy and used only to construct problem-specific grading rubrics, enabling dense process-level rewards that reinforce partial progress and intermediate reasoning steps. On challenging math reasoning benchmarks, ExpRL outperforms SFT, sparse-reward GRPO, and self-distillation as an RL initialization strategy, with additional mixed-domain experiments suggesting broader applicability.
GR2: Generative Reasoning Re-Ranker applies RL with verifiable rewards to industrial-scale LLM recommendation
GR2 (Generative Reasoning Re-Ranker) is a new framework that applies reinforcement learning with verifiable rewards to the re-ranking stage of industrial recommendation systems, a step largely overlooked by prior LLM-based recommendation research. The system combines semantic ID mid-training, reasoning-trace distillation from a stronger teacher model, and purpose-built RL rewards, plus a context compressor and On-Policy Distillation to make it viable at scale. Deployed on industrial traffic, GR2 achieves +18.7% R@1 and +9.6% N@3 over legacy baselines. The paper also identifies a critical reward-hacking failure mode where LLMs exploit position bias or preserve input order, motivating conditional verifiable rewards.
POW3R: Policy-Aware Rubric Rewards for More Efficient RLVR Training
This paper identifies a failure mode in rubric-based reinforcement learning with verifiable rewards (RLVR): static aggregation of criterion weights conflates human-assigned importance with current optimization utility, causing many criteria to be either already saturated or unreachable. The authors introduce POW3R, a framework that dynamically reweights criterion-level rewards during training using rollout-level contrast to emphasize criteria that currently differentiate policy outputs. Across three base policies and two datasets (multimodal and text-only), POW3R wins 24 of 30 comparisons on rubric reward and strict completion metrics, and reaches equivalent performance in 2.5–4× fewer training steps than vanilla GRPO with rubric rewards.
APPO: Fine-grained branching and credit assignment for agentic RL in LLMs
Researchers introduce Agentic Procedural Policy Optimization (APPO), a reinforcement learning method that shifts branching and credit assignment from coarse tool-call boundaries to fine-grained decision points within generated sequences. APPO uses a Branching Score combining token uncertainty with policy-induced likelihood gains to select exploration points, plus procedure-level advantage scaling for credit distribution. Evaluated on 13 benchmarks, APPO improves strong agentic RL baselines by nearly 4 points while maintaining efficient tool use and interpretability. The work addresses a known weakness in multi-turn agentic RL: that influential decisions are distributed throughout sequences, not concentrated at tool-call boundaries.
Reinforcement Learning Recruits a Pre-Existing 'Functional Welfare' Axis in Language Models
Researchers trained language models in a semantically neutral maze environment and extracted concept vectors for rewarded and punished trajectories, finding that RL recruits a pre-existing representational axis encoding functional welfare—how well or badly the system is doing relative to its goals. The punishment vector promotes failure tokens, aligns with negative emotion concepts, and induces refusal and uncertainty when used for steering; the reward vector is its near-antiparallel mirror. Critically, these vectors are effective in models before maze training and appear in pretrain-only models, suggesting the welfare axis pre-exists post-training rather than being created by it. The findings have implications for interpretability, alignment, and understanding how minimal reward signals can broadly reshape model behavior.
AgenticRL: Self-refining LLM-guided reward design and policy refinement for UAV navigation
AgenticRL is a framework that uses a multimodal GPT agent to automate reward function generation, policy training via PPO, and closed-loop self-refinement for UAV navigation tasks. The agent evaluates trained policies through diagnostic feedback, identifies failure modes, and iteratively refines rewards without human intervention. Evaluated across five navigation tasks, the closed-loop refinement improves policy behavior by 71% over initial rewards, with sim-to-real transfer achieving 91% real-world success rate and 94% sim-to-real accuracy.