Semantic Consistency Policy Optimization for Reinforcement Learning of LLM Agents

• URL: http://arxiv.org/abs/2606.25852v1
• Date: Wed, 24 Jun 2026 14:02:13 GMT
• Title: Semantic Consistency Policy Optimization for Reinforcement Learning of LLM Agents
• Abstract summary: Group-based reinforcement learning effectively post-trains agents for long-horizon, sparse-reward tasks by deriving step-level credit from trajectory outcomes.<n>This inconsistency sends conflicting gradients to similar actions and wastes the partially-correct progress inside failed rollouts.<n>We propose Semantic Consistency Policy Optimization ( SCPO), a value-free reward-shaping method that mitigates this inconsistency.<n> SCPO scores each failed step against a successful sibling and adds positive step-level credit for new progress along that sibling.
• Score: 45.62262776712415
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Group-based reinforcement learning effectively post-trains LLM agents for long-horizon, sparse-reward tasks by deriving step-level credit from trajectory outcomes. However, this ties a step's credit to its rollout's final outcome: semantically near-identical intermediate steps receive opposite credit depending on whether their trajectory eventually succeeded or failed. Such semantic credit inconsistency sends conflicting gradients to similar actions and wastes the partially-correct progress inside failed rollouts. Motivated by this, we propose Semantic Consistency Policy Optimization (SCPO), a value-free reward-shaping method that mitigates this inconsistency by recovering step-level credit from successful siblings in the same rollout group. Concretely, SCPO scores each failed step against a successful sibling and adds positive step-level credit for new progress along that sibling. On ALFWorld and WebShop, SCPO matches or exceeds strong group-based baselines, reaching 93.7+/-4.1 percent success on ALFWorld and 74.8+/-2.0 percent on WebShop at 1.5B parameters, with gains concentrated on the hardest multi-step tasks.

Related papers

• PGPO: Potential-Guided Policy Optimization for Multi-Turn Agentic Tasks [41.82007055765541]
  We propose Potential-Guided Policy Optimization for multi-turn agentic tasks.<n>It estimates empirical state potentials from anchor-state-group return statistics within each rollout group.<n>It then derives action advantages from potential differences between adjacent states, enabling cross-trajectory credit propagation.
  arXiv  Detail & Related papers  (2026-09-02T07:44:16Z)
• MileGPO: Milestone Inference with Local Evidence for Graph-Based Policy Optimization of Long-Horizon LLM Agents [26.262385802144323]
  We propose MileGPO, which derives process-level credit from grouped on-policy rollouts through three designs.<n>MileGPO requires neither auxiliary models nor additional environment interaction.
  arXiv  Detail & Related papers  (2026-08-20T08:58:27Z)
• TRCA: Transition-wise Rubric Credit Assignment for Long-horizon LLM Agents [22.01777615106231]
  Long-horizon large language model (LLM) agents are typically optimized with sparse terminal outcomes.<n>We propose Transition-wise Credit Assignment, which derives step-level supervision directly from action-induced transitions.<n>Experiments on ALFWorld, WebShop, and seven search-augmented question-answering benchmarks show consistent improvements over the evaluated baselines.
  arXiv  Detail & Related papers  (2026-08-17T06:19:12Z)
• AgentOPSD: Recursive Self-Distillation for Agentic Reinforcement Learning [58.76655851910778]
  AgentOPSD is a critic-free, recursive method for turn-level credit assignment in agentic reinforcement learning.<n>We evaluate AgentOPSD on ALFWorld, WebShop, and Search-QA using Qwen2.5 models at two scales (3B and 7B)
  arXiv  Detail & Related papers  (2026-08-06T13:00:59Z)
• Hierarchical Advantage Weighting for Online RL Fine-Tuning of VLAs from Sparse Episode Outcomes [40.20394793623452]
  When pretrained VLA policies are fine-tuned through online RL, each rollout episode produces only a single binary outcome.<n>We propose Hierarchical Advantage-Weighted Behavior Cloning (HABC), which trains separate critic heads for viability and efficiency.<n>HABC raises success from supervised fine-tuning (SFT) baselines of 36%, 44%, and 12% to 92%, 88%, and 38%.
  arXiv  Detail & Related papers  (2026-06-15T17:57:14Z)
• When Denser Credit Is Not Enough: Evidence-Calibrated Policy Optimization for Long-Horizon LLM Agent Training [8.367896852036699]
  We propose Evidence-Calibrated Policy Optimization (ECPO), a critic-free policy optimization algorithm that calibrates step-level credit before policy updates.<n>ECPO consistently outperforms strong baselines, improving GiGPO by +5.2/+7.3 success points on ALFWorld/WebShop with Qwen2.5-1.5B while adding only 0.1% additional advantage-computation overhead.
  arXiv  Detail & Related papers  (2026-06-04T08:54:09Z)
• StepOPSD: Step-Aware Online Preference Distillation for Agent Reinforcement Learning [6.365332042924078]
  We present StepOPSD, a post-rollout preference self-distillation framework that takes the agent step as the unit of credit redistribution.<n>StepOPSD decomposes trajectories into action-centered step segments, rescoring them under hindsight-enriched teacher contexts.<n>Results reveal a consistent two-knob law: smaller _clip acts as a broadly stabilizing local trust region, whereas the optimal global mixing strength _mix remains task-dependent.
  arXiv  Detail & Related papers  (2026-05-26T15:07:03Z)
• Milestone-Guided Policy Learning for Long-Horizon Language Agents [67.87164829944238]
  We introduce a milestone-guided policy learning framework, BEACON, to train long-horizon language agents.<n>On ALFWorld, WebShop, and ScienceWorld, BEACON consistently outperforms GRPO and GiGPO.<n>These results establish milestone-anchored credit assignment as an effective paradigm for training long-horizon language agents.
  arXiv  Detail & Related papers  (2026-05-07T12:00:40Z)
• Truncated Step-Level Sampling with Process Rewards for Retrieval-Augmented Reasoning [32.295907409325615]
  Training large language models to reason with search engines via reinforcement learning is hindered by a credit assignment problem.<n>We propose SLATE, a framework built on two complementary ideas.<n> Experiments on seven QA benchmarks confirm that SLATE consistently outperforms both sparse-reward and process-reward baselines.
  arXiv  Detail & Related papers  (2026-02-26T19:05:40Z)
• Verified Critical Step Optimization for LLM Agents [67.05296684575445]
  Critical Step Optimization focuses preference learning on verified critical steps.<n>Method starts from failed policy trajectories rather than expert demonstrations.<n>Experiments on GAIA-Text-103 and XBench-DeepSearch show that CSO achieves 37% and 26% relative improvement over the SFT baseline.
  arXiv  Detail & Related papers  (2026-02-03T11:41:02Z)
• ArenaRL: Scaling RL for Open-Ended Agents via Tournament-based Relative Ranking [84.07076200941474]
  ArenaRL is a reinforcement learning paradigm that shifts from pointwise scalar scoring to intra-group relative ranking.<n>We construct an intra-group adversarial arena and devise a tournament-based ranking scheme to obtain stable advantage signals.<n>Experiments show that ArenaRL substantially outperforms standard RL baselines.
  arXiv  Detail & Related papers  (2026-01-10T08:43:07Z)
• Agentic Reinforcement Learning with Implicit Step Rewards [92.26560379363492]
  Large language models (LLMs) are increasingly developed as autonomous agents using reinforcement learning (agentic RL)<n>We introduce implicit step rewards for agentic RL (iStar), a general credit-assignment strategy that integrates seamlessly with standard RL algorithms.<n>We evaluate our method on three challenging agent benchmarks, including WebShop and VisualSokoban, as well as open-ended social interactions with unverifiable rewards in SOTOPIA.
  arXiv  Detail & Related papers  (2025-09-23T16:15:42Z)
• Group-in-Group Policy Optimization for LLM Agent Training [17.243181792126563]
  Group-in-Group Policy Optimization (GiGPO) is a novel RL algorithm that achieves fine-grained credit assignment for LLM agents.<n>We evaluate GiGPO on challenging agent benchmarks, including ALFWorld and WebShop, as well as tool-integrated reasoning on search-augmented QA tasks.
  arXiv  Detail & Related papers  (2025-05-16T08:26:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.