arXiv subjects

Hanyu Li

Publications and source records attributed to Hanyu Li.

At least 19 recordsLinked to original sources

arXiv2026

PinGER: Token-Efficient Generative Retrieval with Semantic IDs

Generative retrieval replaces nearest-neighbor search with token-by-token decoding of an item's semantic ID (SID). But existing SID-based methods primarily use L levels of tokens sequentially to represent each item in the input history. As a result, a fixed context window holds L times fewer items than it would with one embedding per item. We present PinGER, a generative retriever deployed at Pinterest that retains SID outputs while making input histories compact, built on three ideas. (1) An asymmetric attention mask over interleaved SID and dense tokens lets later items attend only to the dense ones, so the model is supervised at every SID position in training yet serves one token per item while computing the same function. (2) Generation is controllable through goal-conditioned outputs, reinforcement learning, and a prefix-diversity penalty, which together steer retrieval toward surface-specific goals and diverse candidates. (3) A serving stack built for efficient wide-beam generation is paired with decoding constrained to valid identifiers and candidate resolution from a host-local index. Deployed in Related Pins, PinGER lifts saves of fresh Pins by 3.60% and overall saves by 0.62%, while reducing the surface's end-to-end latency by 0.7% and eliminating the cost of the nearest-neighbor search it replaces.

cs.IR↗

arXiv2026

OpenCollab: A Multi-Agent Coding Framework with Programmable Collaboration and Controllable Runtime

Multi-agent coding systems are designed to tackle complex software engineering tasks through collaboration. However, existing evaluations typically assume configured organizations are followed faithfully, whereas reality differs. This behavioral gap, combined with differences in underlying system components, prevents clear attribution of observed gains. To this end, we introduce OpenCollab, a multi-agent coding framework that provides a unified infrastructure for programmable collaboration and controllable runtime. Specifically, OpenCollab unifies organization design, enforces experimental control on a shared runtime, and tracks execution through fine-grained event streams. On this basis, we define Adherence to quantify whether the declared organization is actually realized. Our experiments reveal that agents collaborate very differently across configurations: changing any single dimension shifts Adherence, from 47.2% to as high as 97.2%. Furthermore, extensive agentic coding benchmarks show that OpenCollab (Duo) establishes new SOTA performance compared to the mainstream harnesses such as Mini-SWE-agent, OpenHands, Codex CLI, and Claude Code, showing that a well-designed organization can outperform strong existing harnesses, while OpenCollab's single-agent configuration uses the fewest tokens across all evaluated suites. OpenCollab establishes a unified multi-agent infrastructure for easy programmable collaboration and controlled causal evaluation.

cs.SE↗

arXiv2026

The World Won't Wait: Evaluating and Teaching Agents Under a Live Clock

Large language model (LLM) agents can solve complex interactive tasks, yet reasoning and tool calls take time. In continuously changing environments, an otherwise correct action may arrive too late. We introduce RealtimeBench, with 180 tasks across six game types, to evaluate agents while games continue running during inference. Paired evaluations show that GPT-6.1-sol outperforms the faster Jev when games are paused, but falls behind when they keep running. We propose Two-Timescale Offline Recursive Self-Improvement (TORSI): a slow teacher studies failures and tests corrections in paused games between attempts, then writes reusable instructions for a fast student. The student acts without teacher calls during gameplay, and validation scores determine which updates are adopted. Two rounds raise Jev's average test score from 13.51 to 24.19 on a 0--100 progress scale, improving all six game types. The instructions also improve average scores for three other models without retuning. These findings point toward real-time intelligence: enabling powerful LLM agents to remain effective in a continuously changing world by organizing deliberation, learning, and action across timescales.

cs.AI↗

arXiv2026

MiMo-V2.6: Scaling Reinforcement Learning Towards Self-Improvement

Reinforcement learning (RL) is the central training paradigm for advancing large foundation models towards self-improvement. This report introduces the MiMo-V2.6 series, an omni-modal family that pushes the frontier of model intelligence by scaling RL compute. Prior to RL, we conduct mid-training on a broad multimodal corpus to provide ample exploration space, and build a solid infrastructure on the pretrained hybrid-SWA architecture to support subsequent scale-up. We scale RL compute along three dimensions: (1) larger batches and higher throughput, with an asynchronous training that consumes 1,568 samples and 2.7-3.7B tokens per step at context lengths of up to 1M; (2) more diverse and complex environments, spanning code, general, visual, and cyber domains under a mixture of agent harnesses; and (3) more grader compute, via groupwise agentic grading that yields more accurate reward signals for long-horizon tasks and steers the model towards shorter, more token-efficient solutions. To keep training stable at scale, we freeze the MoE router and establish a multi-layer defense against reward hacking. We further build infrastructure for mixed-task agentic RL, including a unified trajectory representation, high-concurrency multi-framework rollout, decoupled control and data planes, and training-inference consistency. We open-source the training dynamics, RL environments, and RL framework to facilitate reproduction and further research on scaled RL and model self-improvement.

cs.CL↗

arXiv2026

A Simple Algorithm Breaking the 1/3 Barrier for Approximate Nash Equilibria in Bimatrix Games

The PPAD-completeness of computing Nash equilibria in bimatrix games has driven extensive research into polynomial-time algorithms for $ε$-approximate Nash equilibria. Determining the strongest approximation guarantee achievable in polynomial time is a fundamental question in the complexity of approximate Nash equilibria. Despite sustained efforts, breaking the $1/3$ barrier has long been regarded as a difficult open problem. In this paper, we give an affirmative answer by proposing a simple, symmetric deterministic polynomial-time algorithm that has no tunable hyperparameters and achieves an approximation guarantee of approximately $0.30954+δ$. The main idea is to enrich the optimization framework of Tsaknakis and Spirakis with building blocks drawn from our work on strategy mixing and automated approximation analysis, yielding a new algorithmic structure that provides a conceptual explanation for the improved guarantee.

cs.GT↗

arXiv2026

CLOOPD: Closing the Learner Loop in On-Policy Distillation

On-policy distillation (OPD) pays twice for each fresh batch: the student generates trajectories and a stronger teacher scores them. Existing methods improve which trajectories are scored and how the teacher signal is constructed, but usually consume it with one actor update. We introduce CLOOPD, a closed-loop framework separating teacher-signal acquisition from student-side realization. CLOOPD selects an adaptive $α$ waypoint inside a KL envelope, freezes the scored batch and its advantages, re-forwards the student after each actor pass, measures realization, and allocates actor work under a separate token budget. The framework includes deterministic two- and three-pass policies, token-priced CLOOPD-TPMR, and a budget-matched control. Across six 300-step runs on an 8-H20 node, every CLOOPD policy improves the one-pass TOP-D anchor at comparable teacher-token scale: macro accuracy rises from 15.41 to 17.78 with CLOOPD-Fixed2 and 19.36 with CLOOPD-Fixed3. At step 100, CLOOPD-Fixed3 reaches 15.35, nearly matching TOP-D at step 300 while using 67.2% fewer teacher-scored tokens and 28.0% fewer GPU-hours. Earlier 8-A100 ablations show adaptive $α$ eliminates observed trust-envelope violations; a third pass adds headroom. These results position CLOOPD as a framework for budgeting how fully students learn from teacher-scored tokens.

cs.LG↗

arXiv2026

Exact Hill Shares Are Simultaneous Guarantees

Fair division of indivisible bads seeks allocations that guarantee every agent a bundle whose cost is no larger than a meaningful fairness benchmark. The canonical minimax share has widely been used; unfortunately, it is not a simultaneous guarantee. Hill (Ann. Probab., 1987) initiated a complementary approach in which the share depends only on the number of agents and the largest possible single-item value. Li et al. (ACM Trans. Econ. Comput., 2024) gave the exact Hill formula and proved that its monotone closure is a simultaneous guarantee. The closure treats the largest-item cost only as an upper bound and can be strictly larger than the share conditioned on the \emph{actual} largest item. They asked whether this smaller exact share is itself simultaneously guaranteed for three or more agents. We resolve this open question affirmatively. For any number of agents and arbitrary heterogeneous largest-item costs, there is one allocation that satisfies every agent's exact Hill's share. We further provide a polynomial-time algorithm computes such an allocation. Our algorithm combines an ordered moving knife with a tail-domination invariant, and a one-sided trimmed subset-sum routine for the two-agent endpoint without computing an exact minimax partition.

cs.GT↗

arXiv2026

Cardinality-Constrained Randomized Assortments with Balanced Market Share

Assortment optimization asks a seller which of $n$ products to display to each customer. Under the multinomial-logit (MNL) model, a revenue-maximizing policy may concentrate most purchases on only a few products. Balanced market share (BMS) limits this disparity by requiring every product to receive either zero aggregate sales or at least an $α$-fraction of the largest product's sales. We study randomized BMS under a per-assortment cardinality constraint: every displayed assortment contains at most $K$ products, although the seller may rotate a larger catalog across customers. This couples per-realization feasibility with aggregate balance over an exponential policy space. We resolve the model on all three algorithmic fronts. First, reweighting assortments by their MNL denominators turns the cardinality restriction into a single uniform-matroid rank inequality, yielding an exact compact sales-space formulation in which every feasible sales vector can be implemented by a policy supported on $O(n)$ assortments. Second, deciding whether there exists a feasible policy attaining a given revenue target is NP-complete even for $α=1$ and $K=2$, whereas optimization is polynomial-time solvable once the set of products with positive sales is prescribed. Third, despite this hardness, BMS confines all positive sales to one multiplicative band, and a geometric scale search combined with a standard multiple-choice-knapsack dynamic program yields an FPTAS. For every $\varepsilon\in(0,1)$, it achieves at least a $(1-\varepsilon)$ fraction of the optimal expected revenue, preserves BMS and per-assortment cardinality exactly, and outputs a rational policy supported on $O(n)$ assortments.

cs.GT↗

arXiv2026

Approximating Combinatorial Contracts with Arbitrary Costs

We study single-agent combinatorial contracts under linear payments. Under a reward share $α\in[0,1]$, an agent chooses a subset $S$ of $n$ hidden actions, generating reward $f(S)$ at cost $c(S)$, to maximize $αf(S)-c(S)$, while the principal receives $(1-α)f(S)$. For nonnegative additive rewards and monotone supermodular costs, Dütting et al. (SODA 2026) proved an exponential supply-query lower bound for exact optimization and left open whether a representation-independent approximation is possible. We resolve this question in a stronger form, removing all structural assumptions on the cost function: for nonnegative additive rewards, arbitrary normalized nonnegative set costs, and every $\varepsilon\in(0,1)$, we give a deterministic $(1-\varepsilon)$-approximation using $O(n\log(n+1)/\varepsilon)$ supply queries, with query complexity independent of numerical bit lengths and breakpoint separation. More generally, the same guarantee and asymptotic query complexity hold for normalized monotone subadditive rewards and arbitrary normalized nonnegative costs under exact best-response and value-query access. The key idea is a cost-independent, reward-side scale certificate. The algorithm tries all singleton rewards as candidate anchors; one of them brackets the optimal retained share within a factor $n^2$. Geometric search and monotonicity of the induced response reward then yield the approximation without enumerating best-response breakpoints.

cs.GT↗

arXiv2026

SaFeR-Steer: Evolving Multi-Turn MLLMs via Synthetic Bootstrapping and Feedback Dynamics

MLLMs are increasingly deployed in multi-turn settings, where attackers can escalate unsafe intent through the evolving visual-text history and exploit long-context safety decay. Yet safety alignment is still dominated by single-turn data and fixed-template dialogues, leaving a mismatch between training and deployment. To bridge this gap, we propose SaFeR-Steer, a progressive multi-turn alignment framework that combines staged synthetic bootstrapping with tutor-in-the-loop GRPO to train a single student under adaptive, on-policy attacks. We also introduce Trajectory-Consistent Summative Reward (TCSR), which aggregates the historical minimum and average of turn rewards so that any low-quality turn affects the trajectory-level return. I. Dataset. We release STEER, a multi-turn multimodal safety dataset with STEER-SFT (12,934), STEER-RL (2,000), and STEER-Bench (3,227) dialogues spanning 1-10 turns. II. Experiment. Starting from Qwen2.5-VL-3B/7B, SaFeR-Steer substantially improves Safety/Helpfulness on both single-turn (48.30/45.86 $\rightarrow$ 81.84/70.77 for 3B; 56.21/60.32 $\rightarrow$ 87.89/77.40 for 7B) and multi-turn benchmarks (12.55/27.13 $\rightarrow$ 55.58/70.27 for 3B; 24.66/46.48 $\rightarrow$ 64.89/72.35 for 7B), shifting failures to later turns and yielding robustness beyond scaling alone. Code is available at https://github.com/Ed-Bg/SaFeR-Steer-full

cs.LG↗

arXiv2026

ComplexityWorld: Benchmarking Vision-Language Models on Verifiable Visual Decision Making

Vision-language models (VLMs) have made rapid progress in visual perception and increasingly support real-world tasks that depend on images. Many such tasks, however, require more than rec- ognizing what an image contains: a model must use visual evidence to make a complete decision whose parts jointly satisfy global constraints. We introduce COMPLEXITYWORLD, a benchmark of 390 tasks across 39 domain-inspired visual worlds and 29 decision categories. Each task is generated from a hidden structured specification, rendered as a visual scene, and scored by an exe- cutable verifier that accepts any feasible solution. Under direct inference, all evaluated models ex- cept GPT-5.6-Sol remain below 40% verifier ac- ceptance rate (VAR), while GPT-5.6-Sol reaches 75.6%. Performance improves substantially when the same decision information is made explicit in structured form, yet varies sharply across equiva- lent visual presentations. Agent scaffolds provide smaller, model-dependent gains. Together, these results reveal a persistent visual-to-decision bot- tleneck that additional inference alone does not remove.

cs.CV↗

arXiv2026

Statistical equivalence of reduced gravity and enhanced friction in granular packings

Using X-ray tomography, we compare granular packings prepared under buoyancy-reduced effective gravity with normal gravity packings of particles with systematically varied friction. We show that reducing gravity lowers the random loose packing limit in a manner analogous to increasing friction. Granular packings under reduced gravity and with enhanced friction exhibit identical volume distributions, compactivity, and entropy, indicating that both routes sample statistically equivalent Edwards volume ensembles of mechanically stable states. This equivalence originates from a common relaxation of the mechanical stability constraint: under both conditions, fewer particles are required to participate in the underlying load-bearing bridge structures, leading to a lower contact-number requirement and a higher density of mechanically stable states. Nevertheless, reduced gravity retains a distinct contact-scale signature through more isotropic contact orientations. These findings identify gravity as a physical control governing the statistical accessibility of mechanically stable states within the Edwards framework and provide a unified statistical description of granular packings formed through different physical routes.

cond-mat.soft↗

arXiv2026

Discovering Expert-Level Nash Equilibrium Algorithms with Large Language Models

Designing polynomial-time algorithms for approximate Nash equilibria (ANE) with provable worst-case guarantees is a fundamental open problem in algorithmic game theory. While large language models (LLMs) can generate candidate algorithms at scale, certifying worst-case guarantees requires formal analysis over all game instances -- a task for which no automated system previously existed. Here, we present LegoNE, a framework encoding expert proof strategies into a symbolic language that automatically compiles any candidate algorithm into a finite optimization problem certifying its worst-case guarantee. Integrating LegoNE with a reasoning LLM, we rediscovered an algorithm matching the best polynomial-time guarantee for two-player games, and discovered a three-player algorithm improving the best guarantee from $0.6+δ$ to $0.5+δ$ -- provably beyond the reach of the extension technique, the only previously known multi-player ANE design paradigm. These results show that encoding domain-specific proof strategies into a machine-tractable language can support LLM-driven discovery of algorithms outside known human design paradigms.

cs.GT↗

arXiv2026

UniPinRec: Unifying Generative Retrieval and Ranking at Pinterest Scale

Modern recommendation systems predominantly train retrieval and ranking as separate models despite both increasingly relying on large transformers encoding the same user behavior data, duplicating parameters, compute, and serving cost. Prior work unifies the model architecture but not the full pipeline: input formats, training procedures, and serving stacks remain fragmented across stages. We present UniPinRec, which achieves full-stack unification of retrieval and ranking at Pinterest: one input format, one model, one training stage, deployed within existing serving infrastructure. A shared transformer encodes the user action sequence into candidate-independent representations that branch into retrieval (ANN dot-product) and ranking (cross-attention) via task-specific heads. Three ideas make this work: (1) Masked Action Modeling (MAM) eliminates interleaving, enabling weight sharing without doubling context length; (2) Blended training examples pair action sequences with feedview impression slates to satisfy both objectives jointly; (3) Cross-stage KV cache sharing reuses user-history computation from retrieval for ranking, reducing total FLOPs versus serving two independent models. Deployed in the Pinterest core surfaces, UniPinRec delivers approximately +1% online engagement lift while cutting end-to-end serving latency by 11.1% and lifting QPS by 63.6%. To our knowledge, this is the first full-stack unification of retrieval and ranking, covering inputs, model, training and serving, deployed in a production recommendation system.

cs.IR↗

arXiv2026

RepoMirage: Probing Repository Context Reasoning in Code Agents with Perturbations

Code agents are currently having skillful performance on repository-level software engineering benchmarks, but it remains unclear whether success on end-to-end tasks such as issue resolution truly reflects repository context reasoning, the ability to identify the task-relevant information across multiple files and reason over the relations among them. To investigate this question, we introduce RepoMirage, a two-stage evaluation suite built on SWE-Bench Verified that adopts perturbation as a diagnostic tool to increase the demand for context reasoning by transforming how the repository is exposed. First, RepoMirage-Perturb applies three types of semantics-preserving repository-level perturbations, revealing a clear performance drop when correct solving requires broader context access. RepoMirage-Extend further turns perturbation-targeted structural bottlenecks into explicit tasks beyond issue resolution, where the average performance declines from 66.8% in the original setting to 25.3%, indicating a significant deficiency in repository context reasoning. Further trajectory analysis reveals an exploration drift, where agents access broader repository context but fail to turn it into effective structure information. Motivated by this observation, we propose RepoAnchor, a structure-first prototype workflow that separates repository exploration from downstream problem solving, and show that explicit structural scaffolding yields notable gains. These results uncover an previously overlooked gap in repository context reasoning for code agents and suggest that stronger structure-aware methods are potential to improve them.

cs.SE↗

arXiv2026

An Information-Theoretic Criterion for Efficient Data Synthesis

Synthetic data becomes crucial for large language model training, but its effectiveness is highly inconsistent. We provide an information-theoretic account of this inconsistency: synthetic data improves a model only when the generation-training loop is information-open, i.e., shaped by external signals (verifiers, environments, or rubrics) that inject task-relevant information beyond the model's current distribution. When the loop is information-closed (relying on the model's own outputs without such signals), the data processing inequality ensures that task-relevant information can only decrease, making collapse a predicted outcome. Among information-open pipelines, both efficiency and generalization hinge on the meta-level of supervision: a coarser signal such as binary correctness treats all acceptable outputs as equivalent, so the behavior it teaches is not tied to any particular domain or surface form and generalizes naturally across tasks and domains. These observations lead to a guiding thesis: learning preferentially converges to the most information-efficient signal component available, which accelerates learning when that component is the intended one, but causes reward hacking when a spurious pattern happens to be simpler.

cs.LG↗

arXiv2026

A Unified Glassy Rheology for Granular Matter

Granular flows are ubiquitous in nature and industrial applications, yet a complete continuum theory remains a long-standing challenge. The leading empirical approach, μ(I) rheology, lacks microscopic foundations and becomes multivalued in dense, slowly sheared flows where nonlocal corrections are required. Exploiting state-of-the-art high-speed X-ray tomography to investigate microscopic dynamics of dense granular flows in a Couette geometry, we establish a new, universal constitutive law spanning quasi-static to inertial regimes based on structural relaxation, resolving the fundamental difficulty in the original μ(I) framework. By further establishing a non-equilibrium statistical framework for granular flows, we demonstrate an intrinsic analogy between driven granular matter and hard-sphere liquids owing to their identical Carnahan-Starling equation of state, naturally explaining our rheological approach and the emergence of glassy behaviors. Our framework unifies granular rheology with the broader physics of disordered systems and provides a complete, microscopically-based theoretical framework for dense granular flow.

cond-mat.soft↗

arXiv2026

Scalable Preconditioners for the Pseudo-4D DFN Lithium-ion Battery Model

The pseudo-4D Doyle-Fuller-Newman (DFN) model enables predictive simulation of lithium-ion batteries with three-dimensional electrode architectures and particle-scale diffusion, extending the standard pseudo-2D (P2D) formulation to fully resolve cell geometry. This leads to large, nonlinear systems with strong coupling across multiple physical scales, posing significant challenges for scalable numerical solution. We introduce block-structured preconditioning strategies that exploit the mathematical properties of the coupled system, employing multigrid techniques for electrode-level operators and localized solvers for particle-scale diffusion. Comprehensive scalability studies are performed across a range of geometries, including homogeneous and heterogeneous cubic cells, flattened jelly-roll configurations, and triply periodic minimal surface electrodes, to assess solver robustness and parallel scalability. The proposed methods consistently deliver efficient convergence and enable the solution of battery models with hundreds of millions of degrees of freedom on large-scale parallel hardware.

math.NA↗