arXiv subjects

Li Jin

Publications and source records attributed to Li Jin.

At least 19 recordsLinked to original sources

arXiv2026

AgenticSTS: A Bounded-Memory Testbed for Long-Horizon LLM Agents

Memory for a long-horizon LLM agent is a contract about what each future decision is allowed to see. The simplest contract appends past observations, tool calls, and reflections to every prompt, which makes prior context easy to access but also turns it into a jumbled mixture in which the effect of any single memory component is hard to isolate. We introduce and instrument an alternative bounded contract: every decision is made from a fresh user message assembled by typed retrieval, with no raw cross-decision transcript appended. The prompt thus stays bounded across runs of any length, and any single layer can be ablated in isolation. We instantiate the contract in Slay the Spire 2, a closed-rule stochastic deck-building game whose runs require hundreds of tactical and strategic decisions. A public online benchmark of frontier LLMs on the same game reports zero wins at the lowest difficulty across five configurations, and the developer-reported human win rate at the same difficulty is 16%; the task is hard but not saturated. Within our harness, a fixed-A0 ablation shows the largest observed difference when triggered strategic skills are enabled: the no-store baseline wins 3/10 games and adding the skill layer 6/10. At this sample size the comparison is directional rather than statistically decisive (Fisher exact p\approx0.37); a cross-backbone probe and public accumulating-context baselines are reported as operational comparisons rather than controlled tests of the contract variable itself. We release a reproducible testbed: 298 completed trajectories with condition tags, frozen memory/skill snapshots, prompt records, and analysis scripts -- an agent design and a validated, reusable methodology for studying how explicit memory layers shape long-horizon LLM-agent decisions.

cs.AI↗

arXiv2026

Slope-Guided Mamba and Angular-Refined Transformer for Light Field Super-Resolution

Light Field Super-Resolution (LFSR) necessitates accurate modeling of spatial-angular correlations while preserving intrinsic 4D ray coherence. However, maintaining such high-dimensional consistency remains challenging, primarily due to two inherent limitations in prevailing modeling paradigms. First, spatial and angular dimensions are often modeled in a decoupled manner, restricting early cross-dimensional interaction and leading to geometric inconsistencies. Moreover, although continuous sequence modeling paradigms show promise in representing epipolar structures, their rigid scanning mechanisms fundamentally conflict with epipolar geometry, limiting geometry-aware feature aggregation. To address these challenges, we propose a hybrid light field super-resolution network, termed SMART, which integrates a Slope-Guided Mamba and an Angular-Refined Transformer to effectively overcome these limitations. Specifically, we introduce an angular-modulated spatial module to bridge the decoupling gap, incorporating angular priors to strengthen spatial-angular correlation modeling. To mitigate the scan-geometry mismatch, we propose a manifold-aligned trajectory module that enables geometry-consistent sequence modeling along epipolar structures. Experiments on five benchmarks demonstrate that SMART achieves state-of-the-art performance, surpassing previous methods by 0.42 dB (PSNR) with significantly reduced artifacts.

cs.CV↗

arXiv2026

VGPT-RSI for RH-Adjacent Formal Progress: Boundary Certificates, Verified Finite Lagarias Inequalities, and Explicit Failure Localization

The Riemann Hypothesis remains one of the central unsolved problems in mathematics. Rather than claiming proof, we investigate whether a verifiable AI-assisted reasoning system can produce reliable, formally checked partial progress while explicitly identifying the remaining mathematical obstructions. We apply the Verifiable Growing Physical Transformer with Recursive Self-Improvement (VGPT-RSI) to two RH-adjacent certification tasks. First, we construct and verify a finite RH-boundary certificate for inequality on a parameterized safe lower curve over a region. The numerical boundary curve is converted into a certificate-backed lower curve, audited using outward-rounded interval arithmetic and Arb/FLINT ball arithmetic, and then checked in Rocq/CoqInterval for the parameterized theorem. Second, we initiate a formal Lagarias-route certificate. Lagarias criterion states that RH is equivalent to the global inequality. We formalize the finite quantity and produce a Coq-checked finite certificate. The final system identifies the exact unresolved mathematical bottlenecks: formalizing the Lagarias equivalence, proving the global tail theorem beyond any finite cutoff, and potentially reducing counterexamples to colossally abundant or related extremal integers. These results demonstrate that VGPT-RSI can produce certified RH-adjacent formal progress, organize proof dependencies, and avoid overclaiming when the remaining obstruction is genuinely mathematical.

cs.AI↗

arXiv2026

HISR: Hindsight Information Modulated Segmental Process Rewards For Multi-turn Agentic Reinforcement Learning

While large language models excel in diverse domains, their performance on complex longhorizon agentic decision-making tasks remains limited. Most existing methods concentrate on designing effective reward models (RMs) to advance performance via multi-turn reinforcement learning. However, they suffer from delayed propagation in sparse outcome rewards and unreliable credit assignment with potentially overly fine-grained and unfocused turnlevel process rewards. In this paper, we propose (HISR) exploiting Hindsight Information to modulate Segmental process Rewards, which closely aligns rewards with sub-goals and underscores significant segments to enhance the reliability of credit assignment. Specifically, a segment-level process RM is presented to assign rewards for each sub-goal in the task, avoiding excessively granular allocation to turns. To emphasize significant segments in the trajectory, a hindsight model is devised to reflect the preference of performing a certain action after knowing the trajectory outcome. With this characteristic, we design the ratios of sequence likelihoods between hindsight and policy model to measure action importance. The ratios are subsequently employed to aggregate segment importance scores, which in turn modulate segmental process rewards, enhancing credit assignment reliability. Extensive experimental results on three publicly benchmarks demonstrate the validity of our method.

cs.LG↗

arXiv2026

CanoVerse: 3D Object Scalable Canonicalization and Dataset for Generation and Pose

3D learning systems implicitly assume that objects occupy a coherent reference frame. Nonetheless, in practice, every asset arrives with an arbitrary global rotation, and models are left to resolve directional ambiguity on their own. This persistent misalignment suppresses pose-consistent generation, and blocks the emergence of stable directional semantics. To address this issue, we construct \methodName{}, a massive canonical 3D dataset of 320K objects over 1,156 categories -- an order-of-magnitude increase over prior work. At this scale, directional semantics become statistically learnable: Canoverse improves 3D generation stability, enables precise cross-modal 3D shape retrieval, and unlocks zero-shot point-cloud orientation estimation even for out-of-distribution data. This is achieved by a new canonicalization framework that reduces alignment from minutes to seconds per object via compact hypothesis generation and lightweight human discrimination, transforming canonicalization from manual curation into a high-throughput data generation pipeline. The Canoverse dataset will be publicly released upon acceptance. Project page: https://github.com/123321456-gif/Canoverse

cs.CV↗

arXiv2026

CoSMo3D: Open-World Promptable 3D Semantic Part Segmentation through LLM-Guided Canonical Spatial Modeling

Open-world promptable 3D semantic segmentation remains brittle as semantics are inferred in the input sensor coordinates. Yet, humans, in contrast, interpret parts via functional roles in a canonical space -- wings extend laterally, handles protrude to the side, and legs support from below. Psychophysical evidence shows that we mentally rotate objects into canonical frames to reveal these roles. To fill this gap, we propose \methodName{}, which attains canonical space perception by inducing a latent canonical reference frame learned directly from data. By construction, we create a unified canonical dataset through LLM-guided intra- and cross-category alignment, exposing canonical spatial regularities across 200 categories. By induction, we realize canonicality inside the model through a dual-branch architecture with canonical map anchoring and canonical box calibration, collapsing pose variation and symmetry into a stable canonical embedding. This shift from input pose space to canonical embedding yields far more stable and transferable part semantics. Experimental results show that \methodName{} establishes new state of the art in open-world promptable 3D segmentation.

cs.CV↗

arXiv2026

ES-Mem: Event Segmentation-Based Memory for Long-Term Dialogue Agents

Memory is critical for dialogue agents to maintain coherence and enable continuous adaptation in long-term interactions. While existing memory mechanisms offer basic storage and retrieval capabilities, they are hindered by two primary limitations: (1) rigid memory granularity often disrupts semantic integrity, resulting in fragmented and incoherent memory units; (2) prevalent flat retrieval paradigms rely solely on surface-level semantic similarity, neglecting the structural cues of discourse required to navigate and locate specific episodic contexts. To mitigate these limitations, drawing inspiration from Event Segmentation Theory, we propose ES-Mem, a framework incorporating two core components: (1) a dynamic event segmentation module that partitions long-term interactions into semantically coherent events with distinct boundaries; (2) a hierarchical memory architecture that constructs multi-layered memories and leverages boundary semantics to anchor specific episodic memory for precise context localization. Evaluations on two memory benchmarks demonstrate that ES-Mem yields consistent performance gains over baseline methods. Furthermore, the proposed event segmentation module exhibits robust applicability on dialogue segmentation datasets.

cs.CL↗

arXiv2025

LLM-Driven Composite Neural Architecture Search for Multi-Source RL State Encoding

Designing state encoders for reinforcement learning (RL) with multiple information sources -- such as sensor measurements, time-series signals, image observations, and textual instructions -- remains underexplored and often requires manual design. We formalize this challenge as a problem of composite neural architecture search (NAS), where multiple source-specific modules and a fusion module are jointly optimized. Existing NAS methods overlook useful side information from the intermediate outputs of these modules -- such as their representation quality -- limiting sample efficiency in multi-source RL settings. To address this, we propose an LLM-driven NAS pipeline in which the LLM serves as a neural architecture design agent, leveraging language-model priors and intermediate-output signals to guide sample-efficient search for high-performing composite state encoders. On a mixed-autonomy traffic control task, our approach discovers higher-performing architectures with fewer candidate evaluations than traditional NAS baselines and the LLM-based GENIUS framework.

cs.LG↗

arXiv2025

Rectify Evaluation Preference: Improving LLMs' Critique on Math Reasoning via Perplexity-aware Reinforcement Learning

To improve Multi-step Mathematical Reasoning (MsMR) of Large Language Models (LLMs), it is crucial to obtain scalable supervision from the corpus by automatically critiquing mistakes in the reasoning process of MsMR and rendering a final verdict of the problem-solution. Most existing methods rely on crafting high-quality supervised fine-tuning demonstrations for critiquing capability enhancement and pay little attention to delving into the underlying reason for the poor critiquing performance of LLMs. In this paper, we orthogonally quantify and investigate the potential reason -- imbalanced evaluation preference, and conduct a statistical preference analysis. Motivated by the analysis of the reason, a novel perplexity-aware reinforcement learning algorithm is proposed to rectify the evaluation preference, elevating the critiquing capability. Specifically, to probe into LLMs' critiquing characteristics, a One-to-many Problem-Solution (OPS) benchmark is meticulously constructed to quantify the behavior difference of LLMs when evaluating the problem solutions generated by itself and others. Then, to investigate the behavior difference in depth, we conduct a statistical preference analysis oriented on perplexity and find an intriguing phenomenon -- ``LLMs incline to judge solutions with lower perplexity as correct'', which is dubbed as \textit{imbalanced evaluation preference}. To rectify this preference, we regard perplexity as the baton in the algorithm of Group Relative Policy Optimization, supporting the LLMs to explore trajectories that judge lower perplexity as wrong and higher perplexity as correct. Extensive experimental results on our built OPS and existing available critic benchmarks demonstrate the validity of our method.

cs.CL↗

arXiv2025

Probe-and-Release Coordination of Platoons at Highway Bottlenecks with Unknown Parameters

This paper considers coordination of platoons of connected and autonomous vehicles (CAVs) at mixed-autonomy bottlenecks in the face of three practically important factors, viz. time-varying traffic demand, random CAV platoon sizes, and capacity breakdowns. Platoon coordination is essential to smoothen the interaction between CAV platoons and non-CAV traffic. Based on a fluid queuing model, we develop a "probe-and-release" algorithm that simultaneously estimates environmental parameters and coordinates CAV platoons for traffic stabilization. We show that this algorithm ensures bounded estimation errors and bounded traffic queues. The proof builds on a Lyapunov function that jointly penalizes estimation errors and traffic queues and a drift argument for an embedded Markov process. We validate the proposed algorithm in a standard micro-simulation environment and compare against a representative deep reinforcement learning method in terms of control performance and computational efficiency.

eess.SY↗

arXiv2025

From Past To Path: Masked History Learning for Next-Item Prediction in Generative Recommendation

Generative recommendation, which directly generates item identifiers, has emerged as a promising paradigm for recommendation systems. However, its potential is fundamentally constrained by the reliance on purely autoregressive training. This approach focuses solely on predicting the next item while ignoring the rich internal structure of a user's interaction history, thus failing to grasp the underlying intent. To address this limitation, we propose Masked History Learning (MHL), a novel training framework that shifts the objective from simple next-step prediction to deep comprehension of history. MHL augments the standard autoregressive objective with an auxiliary task of reconstructing masked historical items, compelling the model to understand ``why'' an item path is formed from the user's past behaviors, rather than just ``what'' item comes next. We introduce two key contributions to enhance this framework: (1) an entropy-guided masking policy that intelligently targets the most informative historical items for reconstruction, and (2) a curriculum learning scheduler that progressively transitions from history reconstruction to future prediction. Experiments on three public datasets show that our method significantly outperforms state-of-the-art generative models, highlighting that a comprehensive understanding of the past is crucial for accurately predicting a user's future path.

cs.IR↗

arXiv2025

CrossTrace: Efficient Cross-Thread and Cross-Service Span Correlation in Distributed Tracing for Microservices

Distributed tracing has become an essential technique for debugging and troubleshooting modern microservice-based applications, enabling software engineers to detect performance bottlenecks, identify failures, and gain insights into system behavior. However, implementing distributed tracing in large-scale applications remains challenging due to the need for extensive instrumentation. To reduce this burden, zero-code instrumentation solutions, such as those based on eBPF, have emerged, allowing span data to be collected without modifying application code. Despite this promise, span correlation, the process of establishing causal relationships between spans, remains a critical challenge in zero-code approaches. Existing solutions often rely on thread affinity, compromise system security by requiring the kernel integrity mode to be disabled, or incur significant computational overhead due to complex inference algorithms. This paper presents CrossTrace, a practical and efficient distributed tracing solution designed to support the debugging of microservice applications without requiring source code modifications. CrossTrace employs a greedy algorithm to infer intra-service span relationships from delay patterns, eliminating reliance on thread identifiers. For inter-service correlation, CrossTrace embeds span identifiers into TCP packet headers via eBPF, enabling secure and efficient correlation compromising system security policies. Evaluation results show that CrossTrace can correlate thousands of spans within seconds with over 90% accuracy, making it suitable for production deployment and valuable for microservice observability and diagnosis.

cs.NI↗

arXiv2025

Dual-Center Graph Clustering with Neighbor Distribution

Graph clustering is crucial for unraveling intricate data structures, yet it presents significant challenges due to its unsupervised nature. Recently, goal-directed clustering techniques have yielded impressive results, with contrastive learning methods leveraging pseudo-label garnering considerable attention. Nonetheless, pseudo-label as a supervision signal is unreliable and existing goal-directed approaches utilize only features to construct a single-target distribution for single-center optimization, which lead to incomplete and less dependable guidance. In our work, we propose a novel Dual-Center Graph Clustering (DCGC) approach based on neighbor distribution properties, which includes representation learning with neighbor distribution and dual-center optimization. Specifically, we utilize neighbor distribution as a supervision signal to mine hard negative samples in contrastive learning, which is reliable and enhances the effectiveness of representation learning. Furthermore, neighbor distribution center is introduced alongside feature center to jointly construct a dual-target distribution for dual-center optimization. Extensive experiments and analysis demonstrate superior performance and effectiveness of our proposed method.

cs.LG↗

arXiv2025

Learning-Based Cost-Aware Defense of Parallel Server Systems against Malicious Attacks

We consider the cyber-physical security of parallel server systems, which is relevant for a variety of engineering applications such as networking, manufacturing, and transportation. These systems rely on feedback control and may thus be vulnerable to malicious attacks such as denial-of-service, data falsification, and instruction manipulations. In this paper, we develop a learning algorithm that computes a defensive strategy to balance technological cost for defensive actions and performance degradation due to cyber attacks as mentioned above. We consider a zero-sum Markov security game. We develop an approximate minimax-Q learning algorithm that efficiently computes the equilibrium of the game, and thus a cost-aware defensive strategy. The algorithm uses interpretable linear function approximation tailored to the system structure. We show that, under mild assumptions, the algorithm converges with probability one to an approximate Markov perfect equilibrium. We first use a Lyapunov method to address the unbounded temporal-difference error due to the unbounded state space. We then use an ordinary differential equation-based argument to establish convergence. Simulation results demonstrate that our algorithm converges about 50 times faster than a representative neural network-based method, with an insignificant optimality gap between 4\%--8\%, depending on the complexity of the linear approximator and the number of parallel servers.

eess.SY↗

arXiv2025

CatBOX: A Categorical-Continuous Bayesian Optimization with Spectral Mixture Kernels for Accelerated Catalysis Experiments

Identifying optimal catalyst compositions and reaction conditions is central in catalysis research, yet remains challenging due to the vast multidimensional design spaces encompassing both continuous and categorical parameters. In this work, we present CatBOX, a Bayesian Optimization method for accelerated catalytic experimental design that jointly optimizes categorical and continuous experimental parameters. Our approach introduces a novel spectral mixture kernel that combines the inverse Fourier transform of Gaussian and Cauchy mixtures to provide a flexible representation of the continuous parameter space, capturing both smooth and non-smooth variations. Categorical choices, such as catalyst compositions and support types, are navigated via trust regions based on Hamming distance. For performance evaluation, CatBOX was theoretically verified based on information theory and benchmarked on a series of synthetic functions, achieving more than a 3-fold improvement relative to the best-performing baseline and a 19-fold improvement relative to random search on average across tasks. Additionally, three real-life catalytic experiments, including oxidative coupling of methane, urea-selective catalytic reduction, and direct arylation of imidazoles, were further used for in silico benchmarking, where CatBOX reliably identified top catalyst recipes and reaction conditions with the highest efficiencies in the absence of any a priori knowledge. Finally, we develop an open-source, code-free online platform to facilitate trial deployment in real experimental settings, particularly for self-driving laboratory environments.

cs.LG↗

arXiv2025

Embodied World Models Emerge from Navigational Task in Open-Ended Environments

Spatial reasoning in partially observable environments has often been approached through passive predictive models, yet theories of embodied cognition suggest that genuinely useful representations arise only when perception is tightly coupled to action. Here we ask whether a recurrent agent, trained solely by sparse rewards to solve procedurally generated planar mazes, can autonomously internalize metric concepts such as direction, distance and obstacle layout. After training, the agent consistently produces near-optimal paths in unseen mazes, behavior that hints at an underlying spatial model. To probe this possibility, we cast the closed agent-environment loop as a hybrid dynamical system, identify stable limit cycles in its state space, and characterize behavior with a Ridge Representation that embeds whole trajectories into a common metric space. Canonical correlation analysis exposes a robust linear alignment between neural and behavioral manifolds, while targeted perturbations of the most informative neural dimensions sharply degrade navigation performance. Taken together, these dynamical, representational, and causal signatures show that sustained sensorimotor interaction is sufficient for the spontaneous emergence of compact, embodied world models, providing a principled path toward interpretable and transferable navigation policies.

cs.AI↗

arXiv2025

Semi-Gradient SARSA Routing with Theoretical Guarantee on Traffic Stability and Weight Convergence

We consider the traffic control problem of dynamic routing over parallel servers, which arises in a variety of engineering systems such as transportation and data transmission. We propose a semi-gradient, on-policy algorithm that learns an approximate optimal routing policy. The algorithm uses generic basis functions with flexible weights to approximate the value function across the unbounded state space. Consequently, the training process lacks Lipschitz continuity of the gradient, boundedness of the temporal-difference error, and a prior guarantee on ergodicity, which are the standard prerequisites in existing literature on reinforcement learning theory. To address this, we combine a Lyapunov approach and an ordinary differential equation-based method to jointly characterize the behavior of traffic state and approximation weights. Our theoretical analysis proves that the training scheme guarantees traffic state stability and ensures almost surely convergence of the weights to the approximate optimum. We also demonstrate via simulations that our algorithm attains significantly faster convergence than neural network-based methods with an insignificant approximation error.

cs.LG↗

arXiv2025

Benchmarking Multimodal RAG through a Chart-based Document Question-Answering Generation Framework

Multimodal Retrieval-Augmented Generation (MRAG) enhances reasoning capabilities by integrating external knowledge. However, existing benchmarks primarily focus on simple image-text interactions, overlooking complex visual formats like charts that are prevalent in real-world applications. In this work, we introduce a novel task, Chart-based MRAG, to address this limitation. To semi-automatically generate high-quality evaluation samples, we propose CHARt-based document question-answering GEneration (CHARGE), a framework that produces evaluation data through structured keypoint extraction, crossmodal verification, and keypoint-based generation. By combining CHARGE with expert validation, we construct Chart-MRAG Bench, a comprehensive benchmark for chart-based MRAG evaluation, featuring 4,738 question-answering pairs across 8 domains from real-world documents. Our evaluation reveals three critical limitations in current approaches: (1) unified multimodal embedding retrieval methods struggles in chart-based scenarios, (2) even with ground-truth retrieval, state-of-the-art MLLMs achieve only 58.19% Correctness and 73.87% Coverage scores, and (3) MLLMs demonstrate consistent text-over-visual modality bias during Chart-based MRAG reasoning. The CHARGE and Chart-MRAG Bench are released at https://github.com/Nomothings/CHARGE.git.

cs.AI↗