Search arXiv⌕ Search

arXiv subjects

Yuan Cheng

Publications and source records attributed to Yuan Cheng.

At least 19 recordsLinked to original sources
arXiv2026

UniRec: Cross-stage Multi-Task Fusion with Preference Optimization for Cascaded Recommender Systems

Industrial recommender systems use cascaded stages with different candidate spaces, data distributions, and latency constraints. Each stage include multi-objective prediction and multi-task fusion (MTF) modules. Existing MTF methods typically design and optimize the fusion module independently for each stage, which can lead to inconsistent preferences across stages. Although cross-stage coordination methods have been proposed, they mainly transfer supervision to upstream scoring models rather than jointly optimizing the fusion modules. We propose UniRec, a Unified Cross-stage Recommendation Fusion framework that jointly optimizes pre-ranking and ranking fusion agents. UniRec employs a coupled dual-agent architecture with partially shared embeddings and stage-specific fusion functions. It further introduces Dual-Axis Preference Optimization to combine stage-local preferences with downstream-to-upstream preference transfer. To enable stable and efficient cross-stage training, we address two practical challenges: reward-induced distribution drift and the computational cost of pairwise preference learning. Extensive offline and online experiments on a production short-video system demonstrate consistent gains over single-stage fusion and cross-stage coordination baselines. In a nine-week online A/B test, UniRec improves total watch time by $0.675\%$ and app usage time by $0.616\%$, and active users by $0.189\%$, and has been fully deployed in production.

cs.IR↗
arXiv2026

MA-UITD: Offline Multi-Agent Reward Inference for Characterizing Continuous Social Behavior

Multi-agent inverse reinforcement learning offers a powerful framework not only for imitation learning but also for understanding the neural computations supporting social interactions and decision-making. Inferring rewards from interacting agents is challenging because observed choices reflect both personal preferences and expectations about others. Existing methods often require additional environment interaction, repeated policy optimization as an inner loop, or restrictive assumptions about environment dynamics and action spaces, limiting their applicability to fixed behavioral datasets. We introduce Multi-Agent Unilateral Inverse Temporal Difference Learning (MA-UITD), an offline method for inferring explicit agent-specific rewards from joint demonstrations in continuous-state environments with either discrete or bounded continuous actions. Its core idea is to represent the long-term consequences of each agent's own actions through belief-conditioned successor features that average over other agents' behavior. Shared reward features and agent-specific preference vectors connect these unilateral values to the target reward function, while inverse temporal-difference consistency and noise-contrastive behavioral learning enable inference without known dynamics or an inner policy-optimization loop. Across simulated cooperative, competitive, and general-sum tasks, MA-UITD outperforms baselines in discrete-action settings and infers behaviorally compatible reward profiles in continuous-action settings. When applied to experimentally collected social-foraging trajectories, MA-UITD inferred distinct spatial and interaction-dependent reward structures for leaders and followers, providing a quantitative characterization of latent behavioral objectives that may be relevant for understanding the neural computations underlying social behavior.

cs.LG↗
arXiv2026

Router Drift Control: Mitigating Off-Policy Degradation in Mixture-of-Experts Reinforcement Learning

Reinforcement learning (RL) can become off-policy due to training-inference mismatch, mini-batch updates, and asynchronous training. The resulting mismatch is especially harmful in Mixture-of-Experts (MoE) models, where rollout-training discrepancies can change routing decisions and lead to router drift. Rollout Routing Replay (R3) mitigates this mismatch by reusing expert selections from the rollout stage during training. Our analysis shows that routing replay introduces gradient bias by losing contributions from experts selected by the training router but not the rollout router. To address this, we propose Router Drift Control (RDC), which uses routing disagreement between rollout and training as a signal of off-policy staleness and selectively rejects high-drift responses. The sequence-level filtering avoids gradient bias introduced by replay while preserving native expert selection and the associated learning signals for retained responses. Experiments show that RDC achieves larger gains as policy staleness increases and remains effective in asynchronous RL. In asynchronous training, RDC combined with Truncated Importance Sampling (TIS) improves the average peak validation accuracy across AIME 2024 and AIME 2025 by 5.5 points over TIS and 4.5 points over R3.

cs.LG↗
arXiv2026

GOLD: Geometric Optimized Latent Diffusion for Structure-Aware RNA Inverse Folding

Designing RNA sequences that fold into specific 3D structures is a fundamental challenge in therapeutics. Current RNA inverse folding methods typically condition on 3D backbone coordinates but the generation is limited to 1D discrete sequences. This ignores the continuous side-chain orientations, which are essential for overall structural stability. To address this issue, we explicitly introduce side-chain geometry into the RNA generative process. Unlike co-design frameworks that diffuse across separate raw spaces, we propose Geometric Optimized Latent Diffusion (\textbf{GOLD}), which project sequence and structure into a unified latent manifold. GOLD employs a novel GeoVAE to elegantly integrate discrete RNA sequences with continuous side-chain angles. By mapping these modalities into a single, smooth latent manifold, GOLD naturally facilitates a synergistic co-diffusion process. To enhance structural validity, we further introduce a training-free, gradient-guided sampling mechanism. By decoding the generated sequence probabilities and side-chain geometries into structures during the sampling process, we compute analytical gradients from physical constraints. This dynamic feedback actively guides the generated sequence to a physically stable state. Extensive experiments validate the superiority of GOLD, achieving state-of-the-art performance across sequence recovery, secondary structure validity, and 3D foldability metrics. Our code and data are publicly available at https://github.com/darkflash03/GOLD.

cs.AI↗
arXiv2026

VenusRL: A Fully Disaggregated Agentic RL System with Priority Scheduling and Scalable Interaction

Agentic Reinforcement Learning (RL) trains LLM agents through multi-turn interactions with external tool environments. Its multi-turn nature exposes two system-level bottlenecks unaddressed by existing agentic RL frameworks. First, end-to-end training throughput is constrained by the slowest trajectories to complete, yet optimizing per-GPU utilization alone scatters rollout progress across many groups, delaying the completion of enough groups to unblock the next training step. Second, tool sandboxes are statically over-provisioned by their declared memory ceilings, leaving most physical memory stranded while replicating near-identical state across sandboxes launched from the same prompt. We present VenusRL, a fully disaggregated agentic RL system that addresses both bottlenecks. VenusRL's priority-aware action scheduler uses length-prediction heuristics to identify sample groups whose completion is most likely to unblock the next training step, and pushes them ahead of others across batch admission, KV Cache residency, and cross-worker request orchestration. VenusRL's environment resource manager combines a memory-aware admission threshold with a template-keyed page-sharing pool, packing more sandboxes per node while preserving strict memory isolation via write-protected page table entry aliasing and copy-on-write. Across representative agentic RL workloads, VenusRL achieves up to 4.24x end-to-end training speedup over state-of-the-art baselines and reduces environment cost by up to 89%.

cs.DC↗
arXiv2026

AneumoBench: A Source-Linked Benchmark for Synthetic-Geometry Transfer in Aneurysm CFD

Scientific machine learning uses simulation data to train surrogate models for fast physical-field prediction across geometries. Local shape editing can expand limited geometry collections, but whether its variants improve prediction on unseen geometries, and how to allocate them across sources, require controlled evaluation. We introduce AneumoBench, a dataset and benchmark linking 401 source aneurysm geometries to 9,693 locally edited descendant records, with computational fluid dynamics (CFD) fields computed on both. It contains 80,752 steady velocity-pressure cases across eight inlet conditions and 9,715 transient sequences of velocity, pressure, and wall shear stress (WSS). Each sequence contains 100 frames sampled at 0.01-s intervals from a 1-s cardiac cycle. Mesh, point, and voxel interfaces support steady field prediction and WSS forecasting from four observed frames. With family-disjoint splits, we compare source-only training, descendant training, and descendant pretraining followed by source fine-tuning across nine architectures on 79 held-out sources. Under the reported schedules, two-stage training lowers steady-field and reset-window WSS errors relative to source-only training. With the number of sampled fields and training updates fixed within each comparison, GraphSAGE benefits from descendant training and from distributing a fixed number of descendants across more sources. For WSS, reset-window gains do not consistently persist through 96-step rollout, and lower trajectory error need not improve cycle-level shear metrics or hotspot localization. These data and protocols enable researchers to compare descendant selection and training strategies on the same unseen source geometries.

eess.IV↗
arXiv2026

Replay the Curvature: Accurate and Scalable NVFP4 Quantization for Large Language Model Inference

Large language models make weight storage and memory traffic major inference costs, motivating low-precision formats that represent each weight with only a few bits. Such formats use a scale to map floating-point values into a small codebook; NVFP4 improves local range utilization by letting every 16 E2M1 weights share an E4M3 block scale. Choosing that scale is difficult in GPTQ because quantizing one column updates those that follow, so evaluating a block independently can misestimate its final reconstruction error. Large models pose a second challenge: full-precision weights, calibration activations, and second-order state cannot all remain on one accelerator, while assigning complete layers to devices leaves each time-consuming layer solve serial. We introduce \emph{Schur Replay}, a scale-selection algorithm that reproduces the GPTQ updates caused by each block scale and scores the resulting block error after accounting for compensation from unquantized columns. Separately, our execution infrastructure keeps only the active layer resident, tiers activations across device, host, and disk, retires full-precision layers after export, and distributes independent output rows across tensor-parallel ranks. Together, the algorithm and infrastructure attain $99.35\%$ and $100.84\%$ question-weighted recovery from BF16 across seven benchmarks on Qwen3.5-397B-A17B and Llama-3.3-70B-Instruct. On the 397B model, the infrastructure reduces measured per-layer time by $15.17\times$ over ModelOpt and $23.14\times$ over LLM Compressor, with lower memory used per GPU.

cs.LG↗
arXiv2026

Efficient Agentic LLM Serving over SSD-based Sparse KV Storage

Agentic sessions driven by Large language models (LLMs) often alternate between model inference and tool use, accumulating long histories across successive rounds. Serving these sessions efficiently requires reducing attention computation and retaining history key-value (KV) caches to avoid recomputation. Recently, frontier open-source LLMs adopt sparse attention to reduce computation by selecting only part of the history, while SSDs provide a cheaper alternative to CPU DRAM for storing KV caches. However, sparse KV selection depends on the ad hoc intermediate values during model inference, so it forces SSD reads to lie on the inference critical path. These reads are further slowed by fragmented accesses and read-write interference in SSDs. To address these challenges, we present Janus, an agentic serving framework for sparse attention LLMs with SSD-centric KV storage. Janus focuses on append prefill, which processes each round's newly added inputs and accounts for most history KV loading. To move SSD reads out of the critical path, Janus runs the model's own KV selection module on earlier intermediate values, predicting KV demand without additional training. The predicted reads overlap with model computation, and any prediction misses are fetched before attention executes to preserve model outputs. To improve SSD efficiency, Janus coalesces adjacent reads, packs scattered KV pages into sequential writes on the CPU, and limits background writes while reads are active. Across three models and three agentic traces, Janus outperforms existing works by up to 1.57-3.69 times (1.22-1.85 times on average) in terms of the time to first token latency, while maintaining decode efficiency.

cs.DC↗
arXiv2026

Task-Relevant Null-Space Residuals for Non-Injective Neural Mappings

Non-injective mappings in neural networks map distinct inputs to the same representation, thereby implicitly inducing equivalence relations in the input space. However, the input differences eliminated by these mappings may still be required by downstream tasks, creating a mismatch between operator-induced indistinguishability and task-required distinctions. For non-injective linear operators realized in the current forward pass, their null spaces exactly characterize these invisible input variations. We propose Task-Relevant Null-Space Residuals (NSR), a general residual framework for non-injective linear mappings. NSR combines null-space component extraction from pre-mapping representations, member-level encoding and gating, and application-specific integration to exploit potentially task-relevant information under downstream supervision while preserving the original aggregation or merging rules. We evaluate NSR in two structurally different settings: token merging and graph aggregation. In token merging, NSR achieves higher semantic segmentation performance than the corresponding compressed baselines in 34 out of 36 evaluated configurations, with a maximum observed gain of 31.51 mIoU points under strong compression. In graph aggregation, NSR achieves 100% training accuracy on Tree-NeighborsMatch at depths d=2--6 across three backbones, alongside gains on heterophilic node classification and molecular graph regression. Together, these results support null-space residuals as a practical complement to non-injective linear mappings, enabling downstream models to learn from input distinctions invisible in the original operator's output.

cs.AI↗
arXiv2026

Provable Test-Time Scaling for Beam Search in LLM Reasoning

Beam-search-based test-time methods provide an effective way to improve large language model (LLM) performance on long-horizon generation by pruning invalid reasoning paths early, leading to significantly improved reasoning efficiency and more favorable test-time cost scaling. Despite strong empirical success, the theoretical understanding of beam search remains limited. In this paper, we study the test-time compute guarantee of the commonly used beam search framework that uses the model's internal log-likelihood for intermediate scoring, while relying on an external reward model only after a complete response is generated. We first establish a lower bound for vanilla beam search, showing that at least $Ω(C^\star(x)^2)$ samples are required for the optimal response to survive, where $C^\star(x)$ is the token-level coverage coefficient for prompt $x$. This motivates our modified confidence-filtered beam search (CF-Beam), which reduces the sufficient coverage dependence from quadratic to nearly linear under prefix competitiveness, for fixed horizon, gap, and target accuracy. We then show that the regret of CF-Beam is upper-bounded by the probability of rare failure events and the reward estimation error scaled by a path-level coverage coefficient, where the rare-failure term vanishes as per-step sampling increases. Our results highlight a fundamental advantage of beam search over sequence-level inference methods such as Best-of-N and Best-of-Majority. While the guarantees of these approaches typically involve coverage coefficients that grow exponentially with the horizon $L$, CF-Beam controls the dominant search-induced term through a token-level coverage coefficient that scales polynomially with $L$. Our numerical experiments further confirm that beam search is more robust on hard instances and under increasing reasoning horizons.

cs.LG↗
arXiv2026

Finding Kissing Numbers with Game-theoretic Reinforcement Learning

Since Isaac Newton first studied the Kissing Number Problem in 1694, determining the maximal number of non-overlapping spheres around a central sphere has remained a defining challenge in discrete geometry. As the local analogue of Hilbert's 18th problem, it has profound implications across geometry, number theory and information theory. Although lattices and codes have achieved significant progress, the field is confined to isolated extremal configurations, leaving underlying geometric principles obscured. Here we shift the object to the broader extremal configuration space, thereby opening a new path for the Kissing Number Problem. Accordingly, we recast this problem as a cooperative matrix-completion game, and train a reinforcement learning system, PackingStar, to solve it. One player fills cosine entries while the other corrects suboptimal ones, making explosive geometric complexity tractable. Working within extremal configuration spaces, PackingStar discovers new interpretable geometric structures that improve 15 strong bounds held for decades in kissing numbers and their generalizations, several of them provably optimal under natural inner products. These findings reveal the first explicit spherical-code realization of the Fischer group Fi22, extend the classical Euclidean representation of subgroup structure, and directly inspire subsequent breakthroughs by mathematicians. Overall, the work provides an early example of AI-driven progress on a Hilbert-calibre problem, showing how reinforcement learning advances mathematical discovery by unlocking more expressive objects.

cs.LG↗
arXiv2026

DeepSeek Elastic Compute (DSec): A Sandbox Infrastructure for Effective Agentic Training at Scale

Large-scale agentic training and evaluation with large language models (LLMs) rely on isolated, stateful execution environments in which models inspect repositories, invoke tools, execute commands, and interact with task-specific services. These workloads create sandboxes in large bursts, span heterogeneous functionality and isolation requirements, retain state across long interactions, and draw from large image corpora with limited reuse. Supporting them therefore requires an elastic execution platform rather than a single sandbox runtime. This report presents DeepSeek Elastic Compute (DSec), a production sandbox platform that exposes FnCall, container, microVM, and full-VM sandbox backends through a unified SDK. DSec coordinates placement and lifecycle management across the cluster, composes environments from independently versioned layers, combines memory sharing, reclamation, and CPU scheduling for high-density execution, and loads image data on demand from Fire-Flyer File System (3FS), a cluster-wide distributed filesystem. DSec is co-designed with the reinforcement learning (RL) framework, decouples stateful rollout execution from preemptible GPU training, coordinates sandbox lifecycle with training to preserve rollout state while reclaiming idle resources, and mitigates agent misbehavior such as reward hacking. A single production-scale unit of DSec spans around 160 nodes, serving about 3 million sandboxes per day; in production, it supports over 380,000 concurrent sandboxes and sustains over 5,000 sandbox creations per second. Our evaluation and deployment experience show that these mechanisms reduce environment setup and image-distribution overhead, improve memory efficiency, and preserve latency-sensitive performance under high-density overcommit.

cs.DC↗
arXiv2026

MetaStructAtlas: A Grounded 3D Vision-Language Dataset and Benchmark for Functional and Structural Reasoning in Whole-Body PET/CT

The joint interpretation of metabolic function and anatomical structure is essential for clinical diagnosis in whole-body PET/CT. Although recent advances in 3D medical vision-language models have demonstrated remarkable progress, current efforts are limited to regional CT imaging, leaving a critical void in comprehensive whole-body PET/CT analysis. In this work, we introduce MetaStructAtlas, a large-scale dataset for grounded whole-body PET/CT interpretation that synthesizes multimodal imaging with integrated anatomical, metabolic, and semantic annotations. MetaStructAtlas provides 490 co-registered 3D PET and CT volumes with 50,470 organ-level segmentation masks and grounded radiology reports. To facilitate interactive reasoning, we further developed MetaStructVQA, a standardized 3D grounded visual question-answering benchmark containing 100,565 QA pairs. This framework explicitly links diagnostic queries to visual evidence across modalities, encompassing anatomical, morphological, and metabolic characteristics. Finally, we evaluate state-of-the-art 3D medical VLMs on MetaStructVQA, establishing a robust foundation for multimodal representation learning and integrated whole-body reasoning in nuclear medicine.

cs.CV↗
arXiv2026

SemiSAM-O1: Pushing the Boundary of Annotation-Efficient Medical Image Segmentation with Generalist Knowledge Fusion

Semi-supervised learning (SSL) has become a promising solution to alleviate the annotation burden of deep learning-based medical image segmentation models. While recent advances in foundation model-driven SSL have pushed the boundary to extremely limited annotation scenarios, they fail to maintain robust competitive performance in complex imaging modalities. In this paper, we propose SemiSAM-O1, an annotation-efficient framework using only one annotated template case for segmentation. SemiSAM-O1 extends the specialist-generalist collaborative learning framework to the extreme one-label setting by fully exploiting the foundation model's feature representation capability beyond its prompting interface. SemiSAM-O1 operates in two stages. In the first stage, the foundation model's encoder extracts dense features from all volumes, and class prototypes derived from the single annotated template are propagated to the unlabeled pool via feature similarity to produce coarse initial pseudo-labels. In the second stage, an iterative training-and-refinement loop progressively improves both the segmentation model and the pseudo-labels over multiple rounds, where each round trains the model from scratch on current pseudo-labels and generates updated predictions with voxel-wise uncertainty estimates. An uncertainty-guided refinement step further leverages the foundation model's global feature space to correct high-uncertainty regions by aggregating labels from their most similar confident neighbors, establishing a virtuous cycle of mutual improvement. Extensive experiments on a wide range of segmentation tasks across different modalities and anatomical targets demonstrate that SemiSAM-O1 significantly narrows the performance gap between one-label semi-supervised learning and full supervision, while significantly reducing the computational overhead of online foundation model inference.

cs.CV↗
arXiv2026

DualDecoder: Accelerate Long Context LLM Inference by Predictive Prefetch

Long-context inference is becoming a fundamental capability for modern LLM serving, especially driven by emerging agentic applications. Yet it faces a severe memory wall that the KV cache scales proportionally with increasing context length and request concurrency. Existing sparse KV cache methods offload most KV entries to host memory and retrieve only the critical KV entries needed by each decoding step. However, they commonly introduce substantial auxiliary states in GPU memory for KV retrieval management. Our measurements show that these often-overlooked auxiliary states introduce significant memory overhead and become a new bottleneck under high-concurrency workloads. In this paper, we present DualDecoder, a lightweight serving system for long-context LLM inference that enables efficient sparse KV cache retrieval from host memory. Our key insight is that the critical KV entries required for decoding the next token can be accurately predicted from the preceding speculated token. This predictability enables KV retrieval to be proactively prefetched and overlapped with decoding computation, effectively eliminating the GPU memory overhead of auxiliary states. To achieve this prefetching efficiently, DualDecoder leverages a novel dual-token decoding pipeline that accurately identifies critical KV entries with negligible computational overhead, and designs a layer-aware transfer schedule to overlap KV prefetching with model computation and a layer-scoped memory manager to reduce the GPU runtime buffer. Experimental results show that DualDecoder improves decoding throughput by up to 2.62$\times$ over state-of-the-art systems while preserving decoding latency and model quality.

cs.DC↗
arXiv2026

PIVOT: Efficient Query-Group Indexing for Token-Level Sparse Attention

Token-level sparse attention, as implemented by DeepSeek Sparse Attention (DSA) in production systems, makes the downstream attention efficient but shifts the bottleneck to the indexer that feeds it. To select the top-k tokens for each query, the indexer must still score every preceding token, incurring a cost of O(L^2) per layer for a sequence of length L. We observe that this per-query scan is largely redundant: nearby queries select highly overlapping top-k tokens, and the indexer scores are long-tailed along the key axis. We exploit these properties in PIVOT, Proxy Indexing Via One full-prefix Traversal, a training-free, drop-in replacement for the DSA indexer that shares one prefix scan across a group of nearby queries. PIVOT aggregates a group into a single proxy query, performs one shared full-prefix scan to obtain a candidate set, and then selects a top-k for each query from that set. Two variants trade speed for fidelity: PIVOT-Reuse shares the proxy top-k across the group for maximum speed, whereas PIVOT-Refine re-scores the candidate set with the indexer of each query and then selects an individual top-k, matching the dense indexer at a small additional cost. A single algorithm covers both inference phases, differing only in how groups are formed: fixed-size groups of consecutive queries in prefill, and the queries decoded together in one multi-token prediction (MTP) step in decode. On DeepSeek-V3.2 and GLM-5.1 across LongBench and RULER, PIVOT matches the accuracy of the dense DSA indexer while accelerating it by up to 4x and reducing end-to-end latency by up to 1.6x at long context.

cs.CL↗
arXiv2026

A Polynomial Architecture-Attribution Co-Design Framework for Exact Aumann-Shapley Attribution in GNNs

We study feature-level and node-level explanations for graph neural networks (GNNs) through the lens of Aumann-Shapley attribution. Path-integral methods such as Integrated Gradients provide an axiomatic formulation of attribution, but their practical use in deep GNNs typically relies on finite-sample numerical approximations to the path integral, requiring a trade-off between quadrature error and computational cost. This paper proposes APEX, a model-attribution co-design framework that makes the attribution integral exactly computable under a polynomial GNN architecture. The key component is PolyGIN, a GIN-style graph network whose message-passing, normalization, and transformation operations preserve a bounded multivariate polynomial form for scalar model scores, such as pre-softmax logits. We show that, for a PolyGIN with $L$ polynomial transformation blocks, the derivative along the attribution path has degree at most $2^L-1$. Therefore, Gauss--Legendre quadrature can evaluate the Aumann--Shapley path integral exactly, up to floating-point precision, with $2^{L-1}$ deterministic evaluation points. The resulting attributions can be computed at the feature level and then aggregated into node-level scores while preserving completeness. Experiments on synthetic and real-world graph benchmarks show that PolyGIN maintains competitive predictive performance, while the complete APEX framework achieves higher attribution fidelity than the compared baselines and substantially reduces the number of evaluations required for path integration.

cs.LG↗
arXiv2026

Universality Reconsidered: Rethinking the Validation of Foundation Models for General-Purpose 3D Medical Segmentation

Foundation models have emerged as a transformative paradigm in 3D medical imaging, with the promise of unified quantitative analysis across diverse targets and imaging modalities. Yet the prevailing conception of universality remains incomplete. Current models are predominantly developed and evaluated on datasets largely concentrated around a limited set of imaging modalities and anatomical regions. In this Perspective, we evaluate representative 3D segmentation foundation models using paired whole-body structural and functional imaging data. Our analysis reveals a substantial gap between benchmark-reported performance and real-world generalization, with marked degradation on previously unseen data and particularly severe failures on functional imaging modalities. These findings suggest that current foundation models remain far from achieving true universality. We argue that progress requires not only scaling models and datasets, but also a reconsideration of how universality is defined and validated, extending evaluation beyond regional structural benchmarks toward whole-body structural and functional imaging. Our observations highlight the need to distinguish benchmark success from genuine clinical generalization. Bridging this gap will be essential for translating foundation models from controlled evaluation settings to real-world medical practice.

cs.CV↗