mlsanityData to 7 Oct 2026

ARISE: Adapting to Evolving Capability Gaps in Agentic Reinforcement Learning

arXiv 2609.35532 ↗First posted 28 Sep 2026cs.AIcited by 0

On the map: LLM agents and tool use → Reinforcement learning for LLM agents also fits: Agent skill learning and evolution

Abstract (arXiv)

As a long-horizon agent improves through experience, previously observed weaknesses may recede while new limitations emerge, continually changing what it still needs to learn. Yet the learning process often remains tied to a static view of these needs: fixed behavioral criteria and training priorities can become misaligned with evolving agent capabilities, while sparse task-level feedback makes such misalignment more difficult to detect. Even when capability gaps are identified, rollouts from the current policy may repeatedly reproduce the same failures rather than explore better alternatives. To address this, we introduce Adaptive Rubric-Skill Co-Evolution (ARISE), a reinforcement learning framework that uses rollout evidence to continually adapt evaluation criteria, exploration guidance, and training priorities. Rubrics evolve to reward partial behavioral progress, while their paired skills are refined and selectively activated to guide exploration toward unresolved weaknesses. Alongside this co-evolution, capability-based adaptive sampling prioritizes tasks that target behaviors needing further improvement. Experiments on two challenging long-horizon agent benchmarks, SkillsBench and Terminal-Bench, demonstrate that ARISE successfully enhances both overall task performance and training efficiency. The project page is at https://foundation-model-research.github.io/ARISE .

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Reinforcement learning for LLM agents, one of 16 directions in LLM agents and tool use. Builds on 24 of its 26 references that are in the map. First posted on arXiv 28 Sep 2026.

Builds on

24 of 26 references.

Closest ancestors · 2

Same direction, newest first

  1. 2023Voyager: An Open-Ended Embodied Agent with Large Language Models25 May
  2. 2023Reflexion: Language Agents with Verbal Reinforcement Learning20 Mar

Other references · 22

  1. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  2. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  3. 2023Let's Verify Step by Stepcited by 3,598
  4. 2025DAPO: An Open-Source LLM Reinforcement Learning System at Scalecited by 2,678
  5. 2024SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineeringcited by 1,628
  6. 2025gpt-oss-120b & gpt-oss-20b Model Cardcited by 1,258
  7. 2023ExpeL: LLM Agents Are Experiential Learnerscited by 814
  8. 2024Agent Workflow Memorycited by 342
  9. 2025Rubrics as Rewards: Reinforcement Learning Beyond Verifiable Domainscited by 328
  10. 2023Large Language Models as Tool Makerscited by 293
  11. 2026SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Taskscited by 289
  12. 2017Automatic Goal Generation for Reinforcement Learning Agentscited by 252
10 more
  1. 2026SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learningcited by 252
  2. 2025AReaL: A Large-Scale Asynchronous Reinforcement Learning System for Language Reasoningcited by 234
  3. 2020Prioritized Level Replaycited by 190
  4. 2025Checklists Are Better Than Reward Models For Aligning Language Modelscited by 115
  5. 2025DR Tulu: Reinforcement Learning with Evolving Rubrics for Deep Researchcited by 98
  6. 2025Breaking the Exploration Bottleneck: Rubric-Scaffolded Reinforcement Learning for General LLM Reasoningcited by 56
  7. 2025Online Rubrics Elicitation from Pairwise Comparisonscited by 41
  8. 2026Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoningcited by 32
  9. 2026Tmax: A simple recipe for terminal agentscited by 24
  10. 2026KairosAgent: Agentic Time Series Forecasting with Fused Semantic Reasoningcited by 4

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026ARISE-RL: Agentic Rubric-Grounded Iterative Self-Evolution with Reinforcement Learning
  2. 2026ARCO: Adaptive Rubrics with Co-Evolution for Multi-Step LLM-Based Agents
  3. 2026Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning
  4. 2026SkillForge: Co-Evolving Skills and Agents via Dynamic Skill Lifecycles
  5. 2026Rethinking Self-Evolving Agent Skills: Feedback Dynamics over Multiple Rounds
  6. 2026SkillRise: Agentic Reinforcement Learning for Cross-Task Skill Evolution
  7. 2026SkillFocus: Evolving Agent Skills via Capability Decomposition
  8. 2026ARISE: Agent Reasoning with Intrinsic Skill Evolution in Hierarchical Reinforcement Learning