Searcharxiv⌕ Search

arXiv subjects

Shanqi Liu

Publications and source records attributed to Shanqi Liu.

At least 19 recordsLinked to original sources
arXiv2026

From Solver Feedback to Faithful Plans: Multi-Role Reinforcement Learning for Symbolic Planning

Reliable planning requires converting natural-language instructions into executable symbolic specifications, yet large language models remain brittle without costly PDDL annotations and may exploit solver success in semantically unfaithful ways. We study how to learn faithful natural-language-to-PDDL formalization using only solver feedback, without human-written demonstrations. We propose a solvergrounded multi-role reinforcement learning framework where a single language model acts as an Actor, Judge, and Editor for generation, verification, and repair. The Actor proposes PDDL specifications, the Judge provides a solver-calibrated quality signal, and the Editor performs bounded diagnostic-conditioned refinement. On PlanBench, our method improves average success from 35.5% for LLM+P to 70.8%, achieves 66.3% faithful success, and reduces semantic drift to 6.4%. These results show that organizing solver feedback into generation, verification, and repair roles enables more scalable and faithful annotation-free symbolic planning

cs.AI↗
arXiv2026

FlashEvaluator: Expanding Search Space with Parallel Sequence-Level Evaluation

The Generator-Evaluator (G-E) framework generates K candidate sequences and uses an evaluator to select the highest-scoring one, which is widely used in recommender systems (RecSys) and natural language processing (NLP). Existing evaluators commonly score candidates independently. Although such evaluations can be batched, independent scoring neither models interactions among candidates nor eliminates repeated computation of request-level context and recurring candidate elements, causing the total evaluation work to grow approximately linearly with K. To handle with, we propose FlashEvaluator, a joint evaluator that scores all candidate sequences in a single forward pass. FlashEvaluator factorizes evaluation into shared request-level encoding, reusable candidate-side computation, sequence assembly by indexing, and cross-sequence interaction for setwise comparison. We call this request-local reuse scheme QKV-Cache: inspired by autoregressive KV caching, it reuses context-side key/value representations across candidate sequences and, when candidate elements recur, reuses their request-conditioned representations on the query side. In repeated-item settings, the dominant item-encoding cost therefore depends on the number of distinct items rather than their total occurrences across sequences, reducing the marginal cost of evaluating additional candidates. We provide a computational analysis and evaluate FlashEvaluator on recommendation and text summarization. The results show lower latency and higher throughput with competitive recommendation and summarization quality. In an online deployment at Kuaishou with K=50, FlashEvaluator reduces inference latency by 44% and increases QPS by 114% relative to the production baseline, while yielding statistically significant gains in retention, engagement, and ecosystem metrics.

cs.IR↗
arXiv2025

DUET: Dual Model Co-Training for Entire Space CTR Prediction

The pre-ranking stage plays a pivotal role in large-scale recommender systems but faces an intrinsic trade-off between model expressiveness and computational efficiency. Owing to the massive candidate pool and strict latency constraints, industry systems often rely on lightweight two-tower architectures, which are computationally efficient yet limited in estimation capability. As a result, they struggle to capture the complex synergistic and suppressive relationships among candidate items, which are essential for producing contextually coherent and diverse recommendation lists. Moreover, this simplicity further amplifies the Sample Selection Bias (SSB) problem, as coarse-grained models trained on biased exposure data must generalize to a much larger candidate space with distinct distributions. To address these issues, we propose \textbf{DUET} (\textbf{DU}al Model Co-Training for \textbf{E}ntire Space C\textbf{T}R Prediction), a set-wise pre-ranking framework that achieves expressive modeling under tight computational budgets. Instead of scoring items independently, DUET performs set-level prediction over the entire candidate subset in a single forward pass, enabling information-aware interactions among candidates while amortizing the computational cost across the set. Moreover, a dual model co-training mechanism extends supervision to unexposed items via mutual pseudo-label refinement, effectively mitigating SSB. Validated through extensive offline experiments and online A/B testing, DUET consistently outperforms state-of-the-art baselines and achieves improvements across multiple core business metrics. At present, DUET has been fully deployed in Kuaishou and Kuaishou Lite Apps, serving the main traffic for hundreds of millions of users.

cs.IR↗
arXiv2025

MARS: Modality-Aligned Retrieval for Sequence Augmented CTR Prediction

Click-through rate (CTR) prediction serves as a cornerstone of recommender systems. Despite the strong performance of current CTR models based on user behavior modeling, they are still severely limited by interaction sparsity, especially in low-active user scenarios. To address this issue, data augmentation of user behavior is a promising research direction. However, existing data augmentation methods heavily rely on collaborative signals while overlooking the rich multimodal features of items, leading to insufficient modeling of low-active users. To alleviate this problem, we propose a novel framework \textbf{MARS} (\textbf{M}odality-\textbf{A}ligned \textbf{R}etrieval for \textbf{S}equence Augmented CTR Prediction). MARS utilizes a Stein kernel-based approach to align text and image features into a unified and unbiased semantic space to construct multimodal user embeddings. Subsequently, each low-active user's behavior sequence is augmented by retrieving, filtering, and concentrating the most similar behavior sequence of high-active users via multimodal user embeddings. Validated by extensive offline experiments and online A/B tests, our framework MARS consistently outperforms state-of-the-art baselines and achieves substantial growth on core business metrics within Kuaishou~\footnote{https://www.kuaishou.com/}. Consequently, MARS has been successfully deployed, serving the main traffic for hundreds of millions of users. To ensure reproducibility, we provide anonymous access to the implementation code~\footnote{https://github.com/wangshukuan/MARS}.

cs.IR↗
arXiv2024

Limit error distributions of Milstein scheme for stochastic Volterra equations with singular kernels

For stochastic Volterra equations driven by standard Brownian and with singular kernels $K(u)=u^{H-\frac{1}{2}}/\Gamma(H+1/2), H\in (0,1/2)$, it is known that the Milstein scheme has a convergence rate of $n^{-2H}$. In this paper, we show that this rate is optimal. Moreover, we show that the error normalized by $n^{-2H}$ converge stably in law to the (nonzero) solution of a certain linear Volterra equation of random coefficients with the same fractional kernel.

math.PR↗
arXiv2024

Safe and Efficient Online Convex Optimization with Linear Budget Constraints and Partial Feedback

This paper studies online convex optimization with unknown linear budget constraints, where only the gradient information of the objective and the bandit feedback of constraint functions are observed. We propose a safe and efficient Lyapunov-optimization algorithm (SELO) that can achieve an $O(\sqrt{T})$ regret and zero cumulative constraint violation. The result also implies SELO achieves $O(\sqrt{T})$ regret when the budget is hard and not allowed to be violated. The proposed algorithm is computationally efficient as it resembles a primal-dual algorithm where the primal problem is an unconstrained, strongly convex and smooth problem, and the dual problem has a simple gradient-type update. The algorithm and theory are further justified in a simulated application of energy-efficient task processing in distributed data centers.

math.OC↗
arXiv2024

Asymptotic behaviors for Volterra type McKean-Vlasov stochastic integral equations with small noise

This work is devoted to studying asymptotic behaviors for Volterra type McKean-Vlasov stochastic differential equations with small noise. By applying the weak convergence approach, we establish the large and moderate deviation principles. In addition, we obtain the central limit theorem and find the Volterra integral equation satisfied by the limiting process, which involves the Lions derivative of the drift coefficient.

math.PR↗
arXiv2024

True Knowledge Comes from Practice: Aligning LLMs with Embodied Environments via Reinforcement Learning

Despite the impressive performance across numerous tasks, large language models (LLMs) often fail in solving simple decision-making tasks due to the misalignment of the knowledge in LLMs with environments. On the contrary, reinforcement learning (RL) agents learn policies from scratch, which makes them always align with environments but difficult to incorporate prior knowledge for efficient explorations. To narrow the gap, we propose TWOSOME, a novel general online framework that deploys LLMs as decision-making agents to efficiently interact and align with embodied environments via RL without requiring any prepared datasets or prior knowledge of the environments. Firstly, we query the joint probabilities of each valid action with LLMs to form behavior policies. Then, to enhance the stability and robustness of the policies, we propose two normalization methods and summarize four prompt design principles. Finally, we design a novel parameter-efficient training architecture where the actor and critic share one frozen LLM equipped with low-rank adapters (LoRA) updated by PPO. We conduct extensive experiments to evaluate TWOSOME. i) TWOSOME exhibits significantly better sample efficiency and performance compared to the conventional RL method, PPO, and prompt tuning method, SayCan, in both classical decision-making environment, Overcooked, and simulated household environment, VirtualHome. ii) Benefiting from LLMs' open-vocabulary feature, TWOSOME shows superior generalization ability to unseen tasks. iii) Under our framework, there is no significant loss of the LLMs' original ability during online PPO finetuning.

cs.LG↗
arXiv2023

The Onsager-Machlup action functional for degenerate SDEs driven by fractional Brownian motion

In this paper, the explicit expression of Onsager-Machlup action functional to degenerate stochastic differential equations driven by fractional Brownian motion is derived provided the diffusion coeffcient and reference path satisfy some suitable conditions. Then fractional Euler-Lagrange equations for Onsager-Machlup action functional are also obtained. Finally, some examples are provided to illustrate our results.

math.PR↗
arXiv2023

The Onsager-Machlup action functional for degenerate McKean-Vlasov Stochastic Differential Equations

The purpose of this paper is to investigate the existence of the Onsager-Machlup action functional for degenerate McKean-Vlasov stochastic differential equations. To this end, we first derive Onsager-Machlup action functional for degenerate McKean-Vlasov stochastic differential equations with constant diffusion in a broad set of norms by Girsanov transformation, some conditioned exponential inequalities and It$\mathrm{\hat{o}}$ formulas for distribution dependent functional. Then an example is given to illustrate our results.

math.PR↗
arXiv2023

Volterra type McKean-Vlasov SDEs with singular kernels: Well-posedness, Propagation of Chaos and Euler schemes

In this paper, our work is devoted to studying Volterra type McKean-Vlasov stochastic differential equations with singular kernels. Firstly, the well-posedness of Volterra type McKean-Vlasov stochastic differential equations are established. And then propagation of chaos is proved with explicit estimate of the convergence rate. Finally, We also propose an explicit Euler scheme for an interacting particle system associated with the Volterra type McKean-Vlasov equation.

math.PR↗
arXiv2023

Multi-Agent Cooperation via Unsupervised Learning of Joint Intentions

The field of cooperative multi-agent reinforcement learning (MARL) has seen widespread use in addressing complex coordination tasks. While value decomposition methods in MARL have been popular, they have limitations in solving tasks with non-monotonic returns, restricting their general application. Our work highlights the significance of joint intentions in cooperation, which can overcome non-monotonic problems and increase the interpretability of the learning process. To this end, we present a novel MARL method that leverages learnable joint intentions. Our method employs a hierarchical framework consisting of a joint intention policy and a behavior policy to formulate the optimal cooperative policy. The joint intentions are autonomously learned in a latent space through unsupervised learning and enable the method adaptable to different agent configurations. Our results demonstrate significant performance improvements in both the StarCraft micromanagement benchmark and challenging MAgent domains, showcasing the effectiveness of our method in learning meaningful joint intentions.

cs.MA↗
arXiv2023

Controlling Type Confounding in Ad Hoc Teamwork with Instance-wise Teammate Feedback Rectification

Ad hoc teamwork requires an agent to cooperate with unknown teammates without prior coordination. Many works propose to abstract teammate instances into high-level representation of types and then pre-train the best response for each type. However, most of them do not consider the distribution of teammate instances within a type. This could expose the agent to the hidden risk of \emph{type confounding}. In the worst case, the best response for an abstract teammate type could be the worst response for all specific instances of that type. This work addresses the issue from the lens of causal inference. We first theoretically demonstrate that this phenomenon is due to the spurious correlation brought by uncontrolled teammate distribution. Then, we propose our solution, CTCAT, which disentangles such correlation through an instance-wise teammate feedback rectification. This operation reweights the interaction of teammate instances within a shared type to reduce the influence of type confounding. The effect of CTCAT is evaluated in multiple domains, including classic ad hoc teamwork tasks and real-world scenarios. Results show that CTCAT is robust to the influence of type confounding, a practical issue that directly hazards the robustness of our trained agents but was unnoticed in previous works.

cs.MA↗
arXiv2023

Adaptive Value Decomposition with Greedy Marginal Contribution Computation for Cooperative Multi-Agent Reinforcement Learning

Real-world cooperation often requires intensive coordination among agents simultaneously. This task has been extensively studied within the framework of cooperative multi-agent reinforcement learning (MARL), and value decomposition methods are among those cutting-edge solutions. However, traditional methods that learn the value function as a monotonic mixing of per-agent utilities cannot solve the tasks with non-monotonic returns. This hinders their application in generic scenarios. Recent methods tackle this problem from the perspective of implicit credit assignment by learning value functions with complete expressiveness or using additional structures to improve cooperation. However, they are either difficult to learn due to large joint action spaces or insufficient to capture the complicated interactions among agents which are essential to solving tasks with non-monotonic returns. To address these problems, we propose a novel explicit credit assignment method to address the non-monotonic problem. Our method, Adaptive Value decomposition with Greedy Marginal contribution (AVGM), is based on an adaptive value decomposition that learns the cooperative value of a group of dynamically changing agents. We first illustrate that the proposed value decomposition can consider the complicated interactions among agents and is feasible to learn in large-scale scenarios. Then, our method uses a greedy marginal contribution computed from the value decomposition as an individual credit to incentivize agents to learn the optimal cooperative policy. We further extend the module with an action encoder to guarantee the linear time complexity for computing the greedy marginal contribution. Experimental results demonstrate that our method achieves significant performance improvements in several non-monotonic domains.

cs.RO↗
arXiv2022

The Onsager-Machlup action functional for Mckean-Vlasov SDEs

This paper is devoted to deriving the Onsager-Machlup action functional for Mckean-Vlasov stochastic differential equations in a class of norms that dominate $L^2([0,1], \mathbb{R}^d)$, such as supremum norm $\|\cdot\|_{\infty}$, H$\mathrm{\ddot{o}}$lder norms $\|\cdot\|_α$ with $α<\frac{1}{4}$ and $L^p$-norms with $p>4$ are included. Moreover, the corresponding Euler-Lagrange equation for Onsager-Machlup action functional is derived and a example is given.

math.PR↗
arXiv2021

One-shot Face Reenactment Using Appearance Adaptive Normalization

The paper proposes a novel generative adversarial network for one-shot face reenactment, which can animate a single face image to a different pose-and-expression (provided by a driving image) while keeping its original appearance. The core of our network is a novel mechanism called appearance adaptive normalization, which can effectively integrate the appearance information from the input image into our face generator by modulating the feature maps of the generator using the learned adaptive parameters. Furthermore, we specially design a local net to reenact the local facial components (i.e., eyes, nose and mouth) first, which is a much easier task for the network to learn and can in turn provide explicit anchors to guide our face generator to learn the global appearance and pose-and-expression. Extensive quantitative and qualitative experiments demonstrate the significant efficacy of our model compared with prior one-shot methods.

cs.CV↗
arXiv2020

HILONet: Hierarchical Imitation Learning from Non-Aligned Observations

It is challenging learning from demonstrated observation-only trajectories in a non-time-aligned environment because most imitation learning methods aim to imitate experts by following the demonstration step-by-step. However, aligned demonstrations are seldom obtainable in real-world scenarios. In this work, we propose a new imitation learning approach called Hierarchical Imitation Learning from Observation(HILONet), which adopts a hierarchical structure to choose feasible sub-goals from demonstrated observations dynamically. Our method can solve all kinds of tasks by achieving these sub-goals, whether it has a single goal position or not. We also present three different ways to increase sample efficiency in the hierarchical structure. We conduct extensive experiments using several environments. The results show the improvement in both performance and learning efficiency.

cs.LG↗