Searcharxiv⌕ Search

arXiv subjects

Li Chen

Publications and source records attributed to Li Chen.

At least 37 records · Page 2Linked to original sources
arXiv2026

Fewer Paths, Better Performance: Understanding the ZCube Topology through Braess's Paradox

Datacenter networks follow a multipath doctrine: provision many paths between endpoints, hash flows across them, and let redundancy absorb both failures and load imbalance. The ZCube topology violates this doctrine. It removes the Spine layer, eliminates path multiplicity, and cuts one third of switching hardware, yet delivers better performance for both large model training and inference. We explain this anomaly through a structural connection to Braess's paradox, first observed in 1968: both phenomena trace to congestion-oblivious routing over competing paths. Braess showed that adding paths under this condition can hurt; ZCube shows that removing paths under the same condition can help. First, we show that multipath fabrics under structured LLM traffic operate in Braess's shadow: static ECMP hashing is strictly more fragile than greedy routing. Greedy routing reaches an equilibrium within 4/3 of optimal for affine latencies; static hashing admits unbounded imbalance in the worst case. Second, we prove that ZCube is immune to Braess's paradox and that its orthogonal dual partition provably balances load for arbitrary traffic matrices; AllReduce in training and KV cache transfers in disaggregated inference fall out as two corollaries. Third, we quantify the price of this immunity: ZCube trades microsecond hash recovery for millisecond control plane recovery, a trade that upper-layer resilience in LLM serving makes favorable. Production measurements from a cluster serving GLM-5.1 coding inference report 33% lower network cost, 15% higher GPU throughput, and 40.6% lower P99 time to first token. Our analysis suggests that for workloads driven by model structure, matching topology to traffic matters more than path multiplicity.

cs.NI↗
arXiv2026

Perturbation Power Selection for First-Error Delay Maximization in Enhanced SC Decoding

In this paper, we analyze the effect of perturbation power in delaying the first error position, i.e., the first information bit incorrectly decoded by the successive cancellation (SC) decoding. It is conducted over the finite-length perturbation-enhanced SC (PE-SC) decoding paradigm. We show that the FEP delaying probability exhibits a non-monotonic dependence on the perturbation power \(\sigma_{p}^{2}\). Based on this property, an efficient perturbation power selection algorithm that maximizes the delay probability is proposed to enhance the perturbation efficiency. It results in a more efficient perturbation power selection in finite-length PE-SC decoding.

cs.IT↗
arXiv2026

Achievable-Rate Analysis of MISO Systems with Transmit-Side Multiport Matching Networks

Characterizing communication performance under the physical constraints imposed by radio frequency front-end circuits is essential for bridging communication-theoretic analysis and practical circuit design. In this work, we investigate the achievable-rate upper bound of a multiple-input single-output (MISO) system with a transmitter-side interconnected multiport matching network (MMN) under multiport Bode--Fano constraints and develop a rate-oriented MMN circuit realization method. First, based on a circuit-theoretic communication model and the concept of directional gain from multivariable control theory, we reveal how the directional characteristics of MMN power transmission interact with wireless propagation to affect communication performance. Then, building on this directional-gain interpretation, we reformulate the matrix-valued functional optimization problem that characterizes the achievable-rate upper bound and derive the optimal transmission-coefficient structure. Further, a greedy constraint-wise repair method is developed to obtain a feasible suboptimal solution. Finally, through a rate-oriented MMN circuit realization method, we demonstrate that the derived theoretical insights provide effective guidance for practical MMN design. Numerical results validate the theoretical analysis of the achievable-rate upper bound and the effectiveness of the proposed MMN circuit realization method.

eess.SP↗
arXiv2026

UR-VC: Unsupervised Robotic Value Correction for Time-Derived Progress Proxies

Modern robot learning systems increasingly rely on dense progress or value signals to evaluate intermediate states, guide policy learning, and detect task completion, making the quality of these signals critical. Since such dense labels are rarely available at scale, normalized time within a demonstration is often used as a scalable substitute: later frames are treated as higher progress. However, this time-derived label is only a noisy proxy for physical task progress. In contact-rich manipulation, a robot may make progress and then lose it through slips, failed grasps, or partial undoing, while the time-derived label continues to increase monotonically. We introduce Unsupervised Robotic Value Correction (UR-VC), an offline, training-free method for correcting time-derived progress labels. UR-VC exploits a simple regularity in demonstration data: similar states often recur across different episodes, but at different timestamps. Instead of trusting the timestamp from a single trajectory, UR-VC retrieves similar states from other episodes and aggregates their time-derived labels to obtain a corrected progress estimate. UR-VC requires no manual progress labels, reward annotations, or additional value model. We evaluate UR-VC on real bimanual cloth flatten-and-fold data, a long-horizon deformable-object manipulation task with visible intermediate progress. The corrected labels capture local regressions and non-uniform progress that normalized time cannot represent, while preserving the overall task trend. We further use the corrected signal to construct advantage labels for VLA training, following recent advantage-conditioned policy learning. UR-VC shows a positive trend in real-robot task success under matched data, model, and training settings.

cs.RO↗
arXiv2026

Just-In-Time Scene Graph Growth: Combating Perceptual Saturation in Long-Horizon Robotics

While 3D Scene Graphs (3DSGs) provide crucial structured representations for embodied agents, conventional Ahead-of-Time, build-everything-then-filter pipelines conflict with the real-time, low-latency demands of edge platforms, inducing a perceptual saturation effect via severe observation redundancy. To resolve this, we present JITOMA (Just-In-Time On-demand Memory Activation), a closed-loop framework that unifies task reasoning, perception, and memory into a just-in-time growth process. Instead of exhaustively mapping the entire environment, JITOMA leverages a top-down task heatmap at the frontend to filter continuous observations, routing minimal streams to maintain a global foundation of low-cost, dormant anchors. Upon a cognitive query, the backend Large Language Model (LLM) parses the robotic intent to dynamically awaken task-relevant anchors, triggering resource-intensive operations -- such as dense node captioning and functional inference -- exclusively within the activated local subgraph. To evaluate these dynamic capabilities and study perceptual saturation trade-offs, we introduce JITOMA-Bench, a comprehensive suite for long-horizon multi-tasking and complex multi-step reasoning. Extensive experiments demonstrate that JITOMA substantially reduces active graph size and captioning latency, while maintaining stable processing time under long-horizon task switching.

cs.CV↗
arXiv2026

GII-Polar Codes for Block Fading Channels

Polar codes are proven to be capacity-achieving codes, being gradually practiced in wireless communications. However, their successive cancellation (SC) and successive cancellation list (SCL) decoding incur latency challenge especially for long codes. This paper proposes the generalized integrated interleaved (GII)-polar codes for block Rayleigh fading channels, yielding both reduced decoding latency and competent decoding performance. Under the GII paradigm, two consecutive polar codewords of length $N$ are virtually coupled through a nested codebook which has a lower dimension, capable of correcting more errors. The component polar codes are known as interleaves of a GII-polar code. If decoding of an interleave fails, it can be projected into the decoding of the nested one, enabling richer error patterns to be corrected by each interleave. Since decoding of the interleaves can be performed in parallel, GII-polar codes yield a reduced decoding latency over a single polar code of length $2N$. Our simulation results validate both the decoding latency and performance merits of the proposed coding scheme under block Rayleigh fading channel.

cs.IT↗
arXiv2026

LH-AVLN: A Benchmark for Long-Horizon Audio-Visual-Language Navigation

Embodied navigation is moving toward long-horizon missions, yet existing long-horizon benchmarks are largely acoustically silent, and audio-visual navigation tasks typically focus on a single goal. We introduce LH-AVLN, a benchmark for Long-Horizon Audio-Visual-Language Navigation that combines multi-goal mission execution, heterogeneous goal specifications, and persistent spatialized acoustic cues. In LH-AVLN, an agent receives a global mission of two to four goals specified by category, language description, or reference image, and navigates with RGB-D observations, pose, and binaural audio in indoor 3D environments. The benchmark supports both ordered and unordered missions, where alternating goal-associated sounds can guide non-line-of-sight search but may also become distractors as mission progress changes. We further develop PAG-Nav, a training-free reference agent that maintains a temporal uniform semantic map and performs progressive goal-state planning, using sound for search while reserving completion for visual-semantic verification. Experiments show that existing vision-language, memory-based, and audio-visual agents struggle to complete full LH-AVLN missions, and that PAG-Nav provides a stronger diagnostic baseline while leaving substantial room for future progress.

cs.RO↗
arXiv2026

Revisiting the Radial Velocities of Nearby Open Clusters using Gaia DR3

Open clusters (OCs) are essential laboratories for probing stellar dynamics and tracing the structure and evolution of the Milky Way. Accurate measurements of their average radial velocities (RVs) and RV dispersions are crucial for estimating dynamical masses, orbital evolution, and overall kinematic states. Gaia Data Release 3 (DR3) provides an unprecedented volume of high-precision RVs. However, when applied to OCs, Gaia DR3 RVs often yield unusually large and overestimated RV dispersions. This inflation is primarily driven by RV measurement systematics for hot and faint stars, as well as unresolved binary contamination. To mitigate this, we revisit the average RVs and RV dispersions of OCs within 500 pc of the Sun using Gaia DR3. We evaluate the reliability of RV measurements and employ a color-based filtering method. By selecting member stars within an intrinsic color range of $0.2 \le (BP-RP)_{0} \le 1.2$ mag, we exclude hot and cool stars with less reliable RVs. This filtering significantly reduces the inferred RV dispersions while maintaining average RVs consistent with previous literature. Specifically, the median RV dispersion decreases by 26%, dropping from 3.76 km s$^{-1}$ prior to filtering to 2.79 km s$^{-1}$ afterward. These filtered RV dispersions remain systematically larger than tangential velocity dispersions, likely due to residual biases and undetected binaries. However, RV dispersions measured exclusively from red clump giants (found in 6 clusters) are remarkably small ($\lesssim 1.6$ km s$^{-1}$) and align closely with tangential dispersions ($\lesssim 1$ km s$^{-1}$). Ultimately, we provide a practical, Gaia-only strategy to derive more realistic RV dispersions for OCs, identifying red clump giants as exceptionally high-fidelity kinematic tracers for robust cluster studies.

astro-ph.GA↗
arXiv2026

Spin-orbit coupling induced geometric squeezing in rotating Bose-Einstein condensates

Squeezed states play a key role in diverse frontiers of quantum physics. Geometrically squeezed states, a squeezed state in the orbital phase space of rotating Bose-Einstein condensates (BEC), have been conventionally generated by anisotropic trapping potentials. In this work, we propose a different route to generate geometric squeezing via spin-orbit coupling (SOC) in a pseudospin-1/2 BEC. We show that the SOC enables effective two-phonon transitions within the lowest Landau level via virtual spin-flip processes, leading to exponential squeezing dynamics in both spin components. Furthermore, by applying a $\pi/2$ spin rotation, the two spin channels can be coherently coupled to produce two-mode geometric squeezing. We also investigate the influence of interatomic interactions on squeezing performance and identify parameters where robust squeezing can be achieved. Our work provides a viable pathway to realize and manipulate geometric squeezing in spinor quantum gases.

cond-mat.quant-gas↗
arXiv2026

DualEval: Joint Model-Item Calibration for Unified LLM Evaluation

Current LLM evaluation relies on two complementary but often disconnected signals: static benchmarks with objective correctness labels and arena-style preference data that better reflect open-ended user interactions. We introduce DualEval, a latent model-item calibration framework that represents models and evaluation items in a shared space, jointly estimating model ability together with item difficulty and sharpness. We apply DualEval across four domains: coding, math, miscellaneous domain-knowledge tasks, and generic everyday user queries. Our evaluation uses 18 frontier LLMs, static benchmark labels, and reward-model scores validated against held-out human preferences for open-ended model responses. Empirically, our framework produces reliable and balanced model rankings, and its learned item-level profiles support downstream applications such as benchmark compression for sample-efficient evaluation and anomaly detection for contamination or outlier analysis. Overall, DualEval unifies static and arena-style evaluation through joint model-item calibration, producing model rankings and item-level diagnostics that support more sample-efficient, interpretable, and auditable evaluation pipelines.

cs.LG↗
arXiv2026

AI-Assisted Help-Seeking Trajectories in Programming Education from an SRL-Informed Perspective

Generative AI tools provide novice programmers with instant, personalized support, but also raise concerns about whether AI use supports or bypasses students' regulation of problem-solving. Existing work has largely focused on correctness, usability, or overall usage frequency, with less attention to how student--AI help-seeking unfolds. This study addresses this gap by analyzing AI-assisted help-seeking trajectories in university-level programming. Using an SRL-informed analytical framework that links prompt-level help-seeking codes to conceptual, implementation, debugging, and reflective forms of support, we analyzed 1,290 task-specific student prompts linked to 17,190 code submissions from 71 students in introductory Python programming courses. Specifically, we examined how help-seeking interactions were structured across turns and attempts, and how trajectory patterns related to task scores and the number of code submissions. Results indicate that many students primarily used AI for reactive troubleshooting rather than for planned, self-regulated problem-solving. Although trajectory patterns were not associated with significant differences in task scores, they differed substantially in the number of code submissions required. These findings suggest that the educational significance of AI support lies not only in whether students use AI, but in how their help-seeking trajectories develop during programming problem-solving.

cs.AI↗
arXiv2026

World Engine: Towards the Era of Post-Training for Autonomous Driving

Autonomous vehicles must operate safely in the real world, where errors can have severe consequences. Although modern end-to-end driving policies excel in routine scenarios, their reliability is limited by the scarcity of safety-critical ``long-tail'' events in real driving datasets. These rare interactions define the practical safety boundary of the learned policy, yet they are difficult to collect at scale in the real world. Here we show that this fundamental limitation can be addressed by post-training pre-trained driving models on synthesized high-stakes interactions. We introduce World Engine, a generative framework that reconstructs high-fidelity interactive environments from real-world logs and systematically extrapolates them into realistic safety-critical variations. This paradigm enables reinforcement-based post-training to align policies with safety constraints, circumventing the physical risks inherent in real-world exploration. On a public benchmark built on nuPlan, World Engine substantially reduces failures in rare safety-critical scenarios and yields significantly larger gains than scaling pre-training data alone. Furthermore, when deployed on a production-scale autonomous driving system, the resulting policy reduces simulated collisions and demonstrates measurable improvements in on-road testing, showing that post-training on synthesized, safety-critical interactions offers a scalable and effective pathway to safer autonomous driving. The full codebase suite, including training, is released to the public.

cs.RO↗
arXiv2026

Binary Decompilation LLM with Feedback-Driven Multi-Turn Refinement

Binary decompilation is fundamental to security tasks such as vulnerability discovery, malware inspection, and executable-only program understanding. Recent LLM-based decompilation methods have shown promising results, but most still follow a single-turn generation paradigm: given assembly code or decompiler-produced pseudo-code, the model generates one output and stops. Consequently, the generated code may appear readable or even compile successfully, yet still deviate from the behavior of the original binary and mislead downstream analysis. This paper presents AutoDecompiler, a decompilation-specialized LLM trained with reinforcement learning for feedback-driven multi-turn binary decompilation. Instead of treating decompilation as one-shot code generation, AutoDecompiler formulates it as an iterative refinement process, where the model revises generated code based on compilation, execution, and input/output testing feedback. To enable this process, we design decompilation-specific rewards that capture code validity, recompilability, execution consistency, and semantic fidelity. We further construct stage-aware diagnostic feedback from compiler errors, execution failures, and failed test cases, and introduce progress-aware trajectory rewarding and turn-aware advantage reweighting to encourage beneficial revisions while suppressing regressions. We train the AutoDecompiler family and evaluate it across different input settings, model scales, and benchmarks. Experimental results show that AutoDecompiler consistently outperforms its single-turn counterparts under the same model size and input setting, achieving clear improvements in behavioral re-executability. These results demonstrate that learning to exploit program feedback with reinforcement learning is an effective direction for improving the functional correctness of LLM-based binary decompilation.

cs.SE↗
arXiv2026

Martingale Solutions to a Stochastic Keller-Segel System with nonlocal Source and Super-linear Noise

Global nonnegative martingale solutions are shown to exist for a stochastic Keller-Segel system with a nonlocal Fisher-KPP source and super-linear multiplicative noise. The result is obtained for nonnegative initial data with no smallness assumption, provided that the nonlocal source term is dominant. The main difficulty stems from the absence of a coercive structure and the super-linear nature of the noise. An additional cut-off with finite L^2 norm in the classical Galerkin method is added to establish a well-posed approximation problem. Moreover, due to the nonlocal Fisher-KPP structure, it is necessary to prove the positivity of the approximating solution in order to obtain uniform estimates. In the compactness arguments, the usual tightness argument in the framework of Hilbert spaces cannot be directly applied to the uniform estimates obtained in this paper. As a result, we develop a more general version of the compactness argument and tightness criterion, presented in the appendix, which will be applied throughout the paper. This allows for the global existence of nonnegative martingale solutions to be derived from Jakubowski's version of the Skorokhod Theorem, along with a thorough discussion of the convergence properties.

math.PR↗
arXiv2026

RoboNaldo: Accurate, Stable and Powerful Humanoid Soccer Shooting via Motion-Guided Curriculum Reinforcement Learning

Elite humanoid soccer shooting requires whole-body stability, high-impulse whole-body interactions, and accuracy to targets. Motion tracking-driven reinforcement learning (RL) provides stability in whole-body movement coordination, but a fixed reference makes it hard to adapt to varied ball positions and strike timings; in contrast, task reward-driven RL struggles to explore and discover valid kicks from scratch. We therefore introduce RoboNaldo, a three-stage motion-guided curriculum RL framework for high-impulse humanoid interaction. A single human-kick reference is used as a scaffold and progressively shifts optimization towards shooting performance. The curriculum first learns a stable whole-body kicking prior, then adapts the kick to free-kick settings where the ball is stationary at random positions, and finally extends it to moving-ball shooting through a locomotion-command and kick-trigger interface. A high-level heuristic planner controls this interface during training, while alternative high-level controllers can drive the same low-level policy at inference. In simulation, RoboNaldo demonstrates free-kick shot error 48.6% lower and shoot velocity 2.96x than prior work baselines. In real world on a Unitree G1 with onboard perception, RoboNaldo attains 0.73 m and 0.86 m average target shooting error from 3 m away in free-kick and moving-ball cases, accordingly. And the post-contact ball velocity reaches 13.10 m/s, which is 59-71% of reported professional open-play shot speed. Project page: https://opendrivelab.com/RoboNaldo.

cs.RO↗
arXiv2026

Not All Synthetic Data Is Yours to Learn From

Can a language model improve from plain text sampled from itself, with no prompts, no teacher, no verifier, and no reward model? Yes, but only when the synthetic corpus is compatible with the student, a relational property of the source-student pair rather than an intrinsic property of the data. We call this the latent capability resurfacing hypothesis: weak self-training can amplify capabilities already present in the pretrained model, but only under this compatibility condition. We study this in the minimal setting of prompt-free unconditional self-training, where base language models are fine-tuned on text generated from the BOS token alone, with no task specification or external supervision. We report three findings. First, synthetic utility is relational rather than intrinsic: self-generated data is the most effective source, same-lineage transfer outperforms stronger but differently trained sources, and cross-family transfer is substantially weaker. Second, common intrinsic proxies fail: neither benchmark-level semantic similarity nor average per-token likelihood under the student predicts which corpora help. Third, this regime produces a surprising byproduct. In controlled Pythia experiments, capability and verbatim memorization decouple: benchmark utility is preserved or improved while held-out exact-match extraction drops by over 95 percent, with no forget set, privacy objective, or targeted unlearning. Together, these results suggest that prompt-free self-training works by amplifying what the student already knows, not by importing structure from the data. They also reveal a regime in which capability and verbatim memorization can be separated without any explicit unlearning objective.

cs.CL↗
arXiv2026

Configurable Reward Model for Balanced Safety Alignment

Aligning large language models (LLMs) to heterogeneous and rapidly evolving safety requirements remains a critical challenge. Existing instruction-tuned LLMs and standalone safety classifiers often fail to generalize to new safety configurations, motivating the need for Reward Models (RMs) that are explicitly configurable to changing specifications. We introduce the Configurable Safety Reward Model (CSRM), which is jointly optimized for calibrated safety compliance and reward modeling. Our approach is supported by configuration-targeted data augmentation that enforces instruction adherence while preserving relative severity structure. The resulting RM is sensitive to fine-grained safety configurations and conversational nuances, substantially improving generalization to previously unseen safety configurations. CSRM achieves state-of-the-art performance on recent configurable safety benchmarks, including CoSApien (94.6% F1) and DynaBench (75.8% F1), without requiring additional human annotation. When used for downstream safety alignment, CSRM yields LLMs with a significantly improved helpfulness-safety tradeoff compared to existing baselines.

cs.CL↗
arXiv2026

Deconstructing Spatial Complexity: Hierarchical Decomposition for LLM Spatial Reasoning

LLMs have shown remarkable proficiency in general language understanding and reasoning. However, they consistently underperform in spatial reasoning that severely limits their application, particularly in embodied intelligence. Inspired by the success of hierarchical reinforcement learning, this paper introduces a novel method for hierarchical task decomposition in LLM spatial reasoning. Our approach guides LLMs to decompose complex tasks into manageable sub-tasks by identifying key intermediate states and generating simplified sub-environments. However, we identify that LLMs often fail to derive optimal intermediate states due to their insufficient spatial prior, leading to sub-optimal task decomposition. To address this limitation and enhance its planning capability, we propose the MCTS-Guided Group Relative Policy Optimization (M-GRPO), where we reformulate the UCT formula by incorporating the LLM's prior predictive probabilities alongside its epistemic uncertainty. Furthermore, we implement a more fine-grained advantage function, enabling the model to learn optimal path planning. Experimental results demonstrate that our method substantially improves LLM performance on spatial tasks, including navigation, planning, and strategic games, achieving state-of-the-art results. This work paves the way for LLMs in real-world applications.

cs.AI↗