It is found that the relationship between federation and explanation quality is non-monotonic: at near-IID and moderate heterogeneity, federated models match or exceed the faithfulness of a centralized baseline, but at extreme heterogeneity, explanations break down entirely, becoming spatially incoherent and, on averag...
Planning at run time, as in robots that re-plan while they act, needs a planner that reads the PDDL its users write (durative actions, numeric resources, time windows, continuous change), answers quickly, and delivers only validated plans. Specialised planners are strong on individual fragments of PDDL, but they differ...
Francisco Martín Rico, Esther Aguado González, Francisco Miguel Moreno Olivo et al.· 0 citations
Label-free test-time reinforcement learning adapts vision--language models using rewards derived from their own predictions. However, misinterpreted visual evidence and sensitivity to textual formulation can produce a consistent but incorrect consensus that majority voting reinforces. Cross-view aggregation can expose...
Haotian Lyu, Haoyu Wang, Ru Wang et al.· 0 citations
Reliable local search is essential for LLM agents to support everyday decisions and longer plans by finding places that meet users' needs at given times and locations. Yet existing benchmarks provide limited coverage of whether search results satisfy mandatory requirements, account for preferences, and ground factual a...
Keyun Zhang, Dapeng Yin, Fanfu Wang et al.· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
Large language models can often identify useful semantic structure in formal artifacts, but their outputs cannot be trusted as proofs or solver decisions. We study whether LLMs can instead guide formal reasoning systems through a narrow, solver-verifiable interface. We instantiate this idea in quantified array SMT solv...
Kun-Hang Lv, Rui Han, Fuqi Jia et al.· 0 citations
Several corrections of an executable model can fit the same observations while predicting different execution outcomes. We study a repair-to-execution interface that retains compatible finite guard repairs, conditions their weights on diagnostic and ordinary-execution observations, and uses their predictions to choose...
This work introduces AgentDecept, an executable benchmark for evaluating behavioral deception under pressure that enables controlled analysis of when multi-turn agents'reports diverge from execution in realistic application interactions.
Ye-Han Zhang, Qian-Qi Yan, Hang Yin et al.· 0 citations
Skill-evolution methods typically analyze execution trajectories to guide agent skill refinement. However, policy updates can change the behavioral effects and marginal utility of an unchanged skill. Attribution based on earlier trajectories may therefore no longer reflect the skill's contribution under the updated pol...
Yifan Wang, Zhenyi Gai, Yiwen Zhao et al.· 0 citations
Game AI results are comparable only when different implementations expose the same decision problem. Yet seemingly minor representation choices can silently change that problem by altering action probabilities or collapsing states with different outcomes. We study this issue through CoToan, an executable specification...
Whether models trained on different modalities converge to the same representation of the world is central to AI. The Platonic Representation Hypothesis (PRH) conjectures that they do as they scale, but without a proof it offers neither a guarantee nor a recipe. We turn the hypothesis into a theorem. First, we prove a...
Computer-use agents plan on a screenshot for several seconds and then click at a position on the screen; if the page has changed in the meantime, the click lands on an unintended element. Prior work treats the planning interval as the source of this risk. Measuring planned clicks on 31 public web pages, we find that th...
MemSteer is introduced, which turns unlabeled user histories into question-answer (QA) supervision with supporting evidence for memory-design optimization and develops MemSteer-Opt, which combines QA-based design evaluation with evidence-guided failure diagnosis to revise memory store, retrieval, and context constructi...
Jie Huang, Cui-Ling Lan, Qi-Ze Yu et al.· 0 citations
