arXiv AI

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

GraphSkillAA introduces an attribution-guided skill-graph framework that organizes external skills into addressable nodes and typed edges, enabling precise selection, execution, failure attribution, and targeted updates. The system uses a Local Gate to retest affected examples and a Big Gate to evaluate the entire updated graph, rolling back any rejected changes. Experiments with GPT‑5.6‑sol show high performance on SearchQA, LiveMath, and DocVQA, demonstrating that object-level attribution and graph-scoped validation make skill graphs locally optimizable, testable, and reversible.

arXiv AI
Sep 18

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

SkillAA (Skill Abductive Attribution) is a structured skill‑optimization framework designed for frozen language models. It represents skill applicability, execution, and composition in a unified graph, enabling skill selection, attribution‑guided repair, and update validation within the same structure. By contrasting successful and failed executions, SkillAA routes candidate repairs to specific graph objects, updates only the selected local structure, and uses Local and Big Gates to screen candidate changes before commitment, achieving high performance on SearchQA, LiveMath, and DocVQA.

By Ziqiao Shang, Ling-Yue Ge, Lan-Zhe Guo
arXiv Computation and Language
Aug 27

CaSKG: Counterfactual-Causal Skill Graphs for Scalable Agent Skill Retrieval

CaSKG introduces a counterfactual‑causal skill graph framework that calibrates procedural relations before retrieval, building a high‑recall directed candidate graph from semantic, lexical, input/output, and structural evidence and refining it with repair evidence and optional LLM judgment. The framework applies direction‑conditioned textual counterfactual probes—removing, substituting, and reordering skill pairs—to aggregate evidence with Bayesian smoothing, producing a state‑filtered weighted graph for task‑conditioned expansion. Evaluated across six LLM backbones on ALFWorld and ScienceWorld, CaSKG outperforms existing Graph‑of‑Skills methods, improving macro‑average scores and reducing mean environment steps while preserving essential skill dependencies.

By Zhiyuan Li, Linyuan Gao, Xuechun Ding, Hongwei Chen, Yuan Wu, Yi Chang
arXiv Machine Learning
Aug 24

When Graph-JEPA Learns the Wrong Thing: Diagnosing and Repairing Category-Conditional Collapse

The paper investigates a failure mode in Graph-JEPA, a joint‑embedding predictive model trained on a large scientific‑reasoning graph. Despite achieving high linear‑probe accuracy and effective rank, the learned representation contains almost no usable instance information, as shown by retrieval metrics. The authors diagnose the issue to variance allocation in the objective, propose a repair that restores near‑perfect information recovery, and demonstrate that the problem persists even after repair, highlighting limitations in the evaluation metrics used.

By Gollam Rabby, S\"oren Auer
arXiv AI
Jun 3

SkillDAG: Self-Evolving Typed Skill Graphs for LLM Skill Selection at Scale

arXiv:2606. 03056v1 Announce Type: new Abstract: As LLM agents adopt large skill libraries, selecting the right subset becomes a structural problem rather than a similarity-matching one: skills depend on, conflict with, specialize, or duplicate one another, a structure invisible to both full enumeration and embedding similarity.

By Tong Bai, Zhenglin Wan, Pengfei Zhou, Xingrui Yu, Wangbo Zhao, Yang You, Ivor W. Tsang
arXiv AI
Aug 20

Governance Records as Supervision: Verifier-Selected Self-Training for Structured Workflow Repair

The study demonstrates that governance records—structured logs linking task contracts, model attempts, verifier decisions, and outputs—can serve as effective supervision for bounded AI models. Using a verifier-selected self‑training approach, the authors show that a Qwen3‑14B model trained on plans accepted by an independent VAL verifier achieved significant gains in plan acceptance across numerous PlanBench replanning cases, outperforming other selection strategies. The results highlight the feasibility of one‑shot execution and cumulative learning without relying on oracle targets or stronger teachers.

By Jesus Salas
arXiv AI
Jul 24

Workflow-Localized Mechanism Learning: Attribution-Guided Repair and Knowledge Reuse for Structured Agent Skills

arXiv:2607. 20999v1 Announce Type: new Abstract: Agent Skills package reusable procedural knowledge as external artifacts for frozen language-model agents, yet existing optimizers do not jointly resolve where a failure occurs in a workflow, which mechanism caused it, and how relevant knowledge from third-party Skills should be reused locally.

By Zibin Lin, Shengli Zhang, Taotao Wang, Yihan Xia, Deen Ma, Guofu Liao
arXiv AI
Sep 10

Procedural Graphs: Self-Evolving Execution Structures for LLM Agents

The paper introduces Procedural Graphs, a framework that structures procedural knowledge for large language model agents as (procedure, relation, procedure) triplets, analogous to knowledge graphs for factual data. At each decision point, a guidance model uses the local subgraph to bias the agent’s next action, while an LLM refiner self‑evolves the graph by comparing failed and successful trajectories, editing its topology to improve performance. Experiments across various datasets, tasks, and LLMs show that Procedural Graphs consistently outperform memory‑based baselines, and the self‑evolution mechanism further enhances results without manual engineering.

By Yuxing Lu, Yicheng Chen, Shanchan Wu, Sercan \"{O}. Ar{\i}k
arXiv Machine Learning
Sep 4

Selective Hypergraph Refinement for Frozen Graph Clustering

The paper introduces Selective Hypergraph Refinement (SHR), a post‑processing technique for frozen graph clustering models that does not alter model parameters, node representations, or the original graph. SHR uses an attribute hypergraph to generate candidate refinement directions and selectively updates only nodes with sufficient support, preserving the majority of original assignments. Experiments on 15 backbone‑dataset combinations show modest macro gains (up to 0.137 pp) with very few hard assignment changes, indicating a limited but measurable refinement space after training.

By Zimo Si