Xinyuan Song
Publications
Repair the Amplifier, Not the Symptom: Stable World-Model Correction for Agent Rollouts
As agent planning moves from short tool chains toward persistent workflows with thousands or tens of thousands of steps, failures will occur inside large planning graphs rather than in isolated predictions. Replanning the entire graph after every mistake is neither computationally realistic nor desirable: full-graph replay consumes large context budgets, exposes the LLM to many irrelevant symptoms, and can degrade long-context retrieval. This paper studies the missing component in such systems: a world-model corrector that repairs the failed planning graph in place. We compare two families of correctors. The first is the common engineering approach: scan nodes and edges, choose a suspicious local region, and ask an LLM to repair it. We implement strong engineering LLM correctors and find that they can help, especially when given very large contexts. The second family is our approach, WM-SAR (World-Model Subgraph Amplification Repair): instead of scanning for visible symptoms, it works backward from subgraph amplification, identifies the nodes and edges that keep re-amplifying error, and sends only that causal subgraph to the LLM. Across graph simulations and LLM repair experiments, WM-SAR substantially outperforms engineering correctors under realistic token budgets, achieves near-whole-graph stabilization with a compact region, and gives the LLM a cleaner repair target.
Ask the World Before Acting: Budgeted Environment Probing for World-Model Calibration
Long-horizon language agents do not only choose actions; they carry a private model of the world from one decision to the next. When that model drifts, a later failure can be decided before the failing action is ever taken. We study a direct repair mechanism: before committing to the next task action, an agent may ask the environment about one belief field and write the answer back into its world model. This makes environment interaction a scarce calibration resource, not merely a way to advance the task. We introduce \method, a budgeted probing operator for structured belief tables. The useful probes are not the same everywhere. Procedural beliefs, such as tool dependencies, can often be repaired by targeted checks, but those checks spend steps that the task may need. Spatial beliefs, such as object locations and graph edges, rely more on structural cues; the agent's own confidence can be a poor guide when the world changes off-screen. A type-stratified analysis formalizes this probe-action frontier, and controlled experiments show that mid-planning environment evidence reduces terminal world-model error when the probe policy follows the structure of the task.
World-Model Collapse as a Phase Transition
Water looks unchanged as it warms, then at a critical point it boils. We ask whether long-horizon language agents show an analogous transition in their implicit world models. In some parameter settings, changing state load by a small amount, or adding a single step of horizon, leaves behavior nearly unchanged; near a critical boundary, the same small change causes a sudden world collapse. We study this effect in a deterministic task family with exact per-step gold state. A large grid search over state cardinality, dependency density, horizon, branching, observation mode, and mutation rate reveals a phase diagram: a solved plateau, a narrow transition band, and a collapse floor. Per-step traces show the mechanism: world-state fidelity fails before action validity, so the agent is not merely choosing a bad action; it is acting from a corrupted world. Stronger models translate the critical boundary but do not remove the qualitative transition. These results make world-model collapse a measurable bottleneck for long-horizon agents.
AlgoSkill: Learning to Design Algorithms by Scheduling Human-Like Skills
Designing an algorithm from a natural-language problem statement requires identifying the problem structure, reading constraints, choosing a suitable paradigm, checking correctness, and refining complexity. Existing large language model (LLM) methods often rely on direct generation or generic self-refinement, leaving these steps implicit. We propose AlgoSkill, which models algorithm design as sequential decision-making over a typed library of algorithmic skills, including abstraction, constraint analysis, state design, data-structure selection, proof checking, counterexample construction, and complexity refinement. A learned scheduler proposes skills from the current design state, while a Monte Carlo Tree Search (MCTS) controller explores skill sequences using verification feedback from compilation, testing, stress testing, and complexity analysis. Experiments on competitive programming and combinatorial optimization benchmarks show that AlgoSkill improves over direct LLM generation, chain-of-thought prompting, self-refinement, and MCTS without typed skills. Ablations show that typed skills, verification-based repair, and search-based scheduling each contribute to performance. These results support treating automatic algorithm design as verification-guided skill scheduling rather than one-shot code generation.
Manifold-Aware Temporal Domain Generalization for Large Language Models
Temporal distribution shifts are pervasive in real-world deployments of Large Language Models (LLMs), where data evolves continuously over time. While Temporal Domain Generalization (TDG) seeks to model such structured evolution, existing approaches characterize model adaptation in the full parameter space. This formulation becomes computationally infeasible for modern LLMs. This paper introduces a geometric reformulation of TDG under parameter-efficient fine-tuning. We establish that the low-dimensional temporal structure underlying model evolution can be preserved under parameter-efficient reparameterization, enabling temporal modeling without operating in the ambient parameter space. Building on this principle, we propose Manifold-aware Temporal LoRA (MaT-LoRA), which constrains temporal updates to a shared low-dimensional manifold within a low-rank adaptation subspace, and models its evolution through a structured temporal core. This reparameterization dramatically reduces temporal modeling complexity while retaining expressive power. Extensive experiments on synthetic and real-world datasets, including scientific documents, news publishers, and review ratings, demonstrate that MaT-LoRA achieves superior temporal generalization performance with practical scalability for LLMs.
Learning to Recommend Multi-Agent Subgraphs from Calling Trees
Multi-agent systems (MAS) increasingly solve complex tasks by orchestrating agents and tools selected from rapidly growing marketplaces. As these marketplaces expand, many candidates become functionally overlapping, making selection not just a retrieval problem: beyond filtering relevant agents, an orchestrator must choose options that are reliable, compatible with the current execution context, and able to cooperate with other selected agents. Existing recommender systems -- largely built for item-level ranking from flat user-item logs -- do not directly address the structured, sequential, and interaction-dependent nature of agent orchestration. We address this gap by \textbf{formulating agent recommendation in MAS as a constrained decision problem} and introducing a generic \textbf{constrained recommendation framework} that first uses retrieval to build a compact candidate set conditioned on the current subtask and context, and then performs \textbf{utility optimization} within this feasible set using a learned scorer that accounts for relevance, reliability, and interaction effects. We ground both the formulation and learning signals in \textbf{historical calling trees}, which capture the execution structure of MAS (parent-child calls, branching dependencies, and local cooperation patterns) beyond what flat logs provide. The framework supports two complementary settings: \textbf{agent-level recommendation} (select the next agent/tool) and \textbf{system-level recommendation} (select a small, connected agent team/subgraph for coordinated execution). To enable systematic evaluation, we construct a unified calling-tree benchmark by normalizing invocation logs from eight heterogeneous multi-agent corpora into a shared structured representation.