mlsanityData to 14 Oct 2026

GraphSkillAA: Attribution-Guided Skill-Graph Updating with Targeted Validation and Rollback

arXiv 2609.20455 ↗First posted 17 Sep 2026cs.AIcited by 1

Not on the map. Why?

Abstract (arXiv)

External skills provide domain knowledge and procedures without updating model parameters, but flat collections obscure skill applicability, dependencies, and composition. Graphs organize skills into addressable nodes and explicit relations, supporting selection and composition. Yet existing skill-graph methods use this structure mainly for retrieval, rather than to localize updates, scope retesting, or precisely roll back rejected changes. We introduce GRAPHSKILLAA (GraphSkill Abductive Attribution), which uses one addressable graph for skill selection, execution, failure attribution, targeted updating, validation, and rollback. Nodes separate applicability, execution, and exclusion conditions; typed edges encode prerequisite and enhancement relations. The frozen student records used nodes and edges, while the teacher contrasts related successes and failures to route each supported repair to the smallest relevant field or relation; execution lapses or insufficient evidence leave the graph unchanged. A Local Gate retests affected examples, while a Big Gate evaluates the merged graph on the complete update pool; rejected changes are rolled back. With GPT-5.6-sol, GRAPHSKILLAA reaches 81.5%, 66.7%, and 91.2% on SearchQA, LiveMath, and DocVQA, respectively, and attains the highest observed mean in every main setting. These results show that object-level attribution and graph-scoped validation make a skill graph a locally optimizable, testable, and reversible external state.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 26 of its 37 references that are in the map. First posted on arXiv 17 Sep 2026.

Builds on

26 of 37 references.

References · 26

  1. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,182
  2. 2023Voyager: An Open-Ended Embodied Agent with Large Language Modelscited by 2,191
  3. 2023ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIscited by 1,947
  4. 2024From Local to Global: A Graph RAG Approach to Query-Focused Summarizationcited by 1,453
  5. 2023Gorilla: Large Language Model Connected with Massive APIscited by 1,332
  6. 2020DocVQA: A Dataset for VQA on Document Imagescited by 1,317
  7. 2020ALFWorld: Aligning Text and Embodied Environments for Interactive Learningcited by 1,205
  8. 2023Large Language Modelscited by 780
  9. 2023Automatic Prompt Optimization with "Gradient Descent" and Beam Searchcited by 653
  10. 2026SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Taskscited by 315
  11. 2026SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learningcited by 278
  12. 2017SearchQA: A New Q&A Dataset Augmented with Context from a Search Enginecited by 225
14 more
  1. 2025SkillWeaver: Web Agents can Self-Improve by Discovering and Honing Skillscited by 192
  2. 2026Trace2Skill: Distill Trajectory-Local Lessons into Transferable Agent Skillscited by 187
  3. 2026MemSkill: Learning and Evolving Memory Skills for Self-Evolving Agentscited by 142
  4. 2023CRAFT: Customizing LLMs by Creating and Retrieving from Specialized Toolsetscited by 129
  5. 2026CoEvoSkills: Self-Evolving Agent Skills via Co-Evolutionary Verificationcited by 128
  6. 2026SkillRouter: Skill Routing for LLM Agents at Scalecited by 61
  7. 2023ControlLLM: Augment Language Models with Tools by Searching on Graphscited by 59
  8. 2026Graph-of-Skills: Dependency-Aware Structural Retrieval for Massive Agent Skillscited by 55
  9. 2024ToolRerank: Adaptive and Hierarchy-Aware Reranking for Tool Retrievalcited by 41
  10. 2026SkillDAG: Self-Evolving Typed Skill Graphs for LLM Skill Selection at Scalecited by 16
  11. 2026SkillGraph: Skill-Augmented Reinforcement Learning for Agents via Evolving Skill Graphscited by 15
  12. 2026Demystifying Agent Skills: Why They Work-Until They Don'tcited by 5
  13. 2026CaSKG: Counterfactual-Causal Skill Graphs for Scalable Agent Skill Retrievalcited by 3
  14. 2026When Not to Imitate: Boundary-Aware Skill Memory for Reliable Tool-Use LLM Agentscited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026From Existing Methods to New Research Ideas: Two Perspectives on Recursive Self-Improvementcited by 0