Peer Revenue An exchange of ideas

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Tuesday 6 October

Machine Learning

TRACE: Rollout-Guided Quantization-Aware Training for FP4 Reinforcement Learning of MoE Language Models

Xin Wang, Hao Yu, Zhengyang Zhuge et al. (+9 more)

Reinforcement learning (RL) for post-training large language models (LLMs) incurs substantial computation and memory overhead during rollout generation, which motivates low-precision rollout for efficient RL training. However, existing FP4 RL methods suffer from a key limitation: they primarily optimize quantization accuracy on the training and rollout paths independently rather than directly reducing the discrepancy between the two quantized execution paths. In this work, we propose TRACE (Train-Rollout Quantization Alignment via Compact GuidancE), an FP4 quantization framework for RL training of Mixture-of-Experts (MoE) language models that addresses the limitation of existing FP4 RL methods. TRACE incorporates rollout-guided quantization-aware training that uses rollout-side quantization outcomes to guide training-side FP4 rounding decisions, directly reducing train-rollout discrepancy. Moreover, TRACE adopts an efficient quantization-information caching scheme that selectively retains mantissa and scale information from deeper layers to reduce the storage and communication overhead introduced by rollout guidance. We evaluate TRACE on four large-scale MoE language models across reasoning, coding, and long-horizon RL tasks. Our results demonstrate that TRACE enables joint FP4 weight/activation and FP4 KV-cache rollout with RL performance comparable to BF16 rollout, while achieving up to 5.4xrollout speedup and strong final FP4 performance compared with post-hoc FP4 quantization of BF16-trained policies.
Read the full abstract

All authors Xin Wang, Hao Yu, Zhengyang Zhuge, Bochao Mao, Zheng Li, Junda Feng, Yuyan Luo, Yi Zhang, Yizhong Cao, Mi Zhang, Dayiheng Liu, Jianwei Zhang

Artificial Intelligence

OTel: Open Telco AI Datasets, Benchmarks, and Models

Farbod Tavakkoli, Gregory Diamos, Kenneth Church et al. (+14 more)

We present Open Telco (OTel), an open telecom AI resource that releases derived telecom datasets for retrieval, reranking, instruction tuning, and safety/abstention, together with 30 full-parameter post-trained baselines spanning 10 embedding models, 3 rerankers, and 17 language models. The community has already engaged substantially with the resource: as of May 3, 2026, the released models have been downloaded over 16 million times and the project has received 157+ pieces of media coverage worldwide. Building on prior open telecom datasets and benchmarks, OTel provides documented telecom data sources, held-out evaluation partitions, trained embedding models, rerankers, context-grounded LLMs, and safety/abstention data in one unified resource. Each baseline starts from an open-weight model and is post-trained on OTel-derived data using an open training recipe, then evaluated on held-out OTel evaluation partitions. OTel post-training improves performance across all three model families: embedding retrieval reaches 93.1% NDCG@10, reranking reaches 0.947 MRR@10, and language-model correctness reaches 87.8%. We release OTel as a reproducible starting point and invite the community to expand the data, improve embedding and reranking models, and build stronger context-grounded telecom LLMs.
Read the full abstract

All authors Farbod Tavakkoli, Gregory Diamos, Kenneth Church, Roderic Paulk, Jorden Terrazas, David Kanter, Mark Austin, Imtiaz Karim, Syed Ali Raza Zaidi, Lina Bariah, Merouane Abdelkader Debbah, Ali Maatouk, Louis Powell, Nikolaos Vasiloglou, Ashish Vaswani, Somanshu Singla, Adarsh Chaluvaraju

Number Theory

The optimal van der Corput property for squares

Steve Fan and Andrew Lott

We prove a sharp quantitative form of the van der Corput property for squares. For all sufficiently large $N$, we construct a real trigonometric polynomial $P(\theta)=a_0+\sum_{1\le n\le\sqrt N}a_n\cos(2\pi n^2\theta)$ satisfying \begin{equation}\tag{1}\label{eq:abstract-vdc} P(\theta)\ge0\quad(\theta\in\mathbb{R}/\mathbb{Z}),\qquad P(0)=1, \end{equation} with \[0\le a_0\le\exp(-c\sqrt{\log N\log\log N}),\] where $c>0$ is an absolute constant. As a consequence, if $A\subseteq\mathbb{N}\cap[1,N]$ has no nonzero square difference, then \[|A|\le N\exp(-c\sqrt{\log N\log\log N}),\] improving upon a recent result of Green and Sawhney. Our construction builds on previous work of Slijepčević, Slijepčević–Ninčević, and Fan–Lott, and the proof of the upper bound is elementary and self-contained. We also show that every real trigonometric polynomial of the above form satisfying \eqref{eq:abstract-vdc} has \[a_0\ge\exp(-C\sqrt{\log N\log\log N})\] for an absolute constant $C>0$. Thus the van der Corput property cannot be used to obtain a power-saving bound for the Furstenberg–Sárközy theorem.
Read the full abstract

Computation and Language

No Transformer Beats Six Covariates: Long-Horizon Prediction of Depressive Symptoms from Childhood Essays

Daniel Kua, Emrul Hasan, John-Jose Nunez et al. (+1 more)

Natural language processing (NLP) models can detect depression-related language in text written near the time symptoms are measured, but whether pretrained transformers can predict depressive symptoms from text written twelve years earlier is largely untested. In the National Child Development Study, a British birth cohort, we predict probable depressive symptoms at age 23 from essays the same people wrote at age 11. Our baseline, a logistic regression on six childhood covariates, outperforms every text model that sees only the essay: seven fine-tuned transformers, a bag-of-words model, frozen embeddings and four zero-shot large language models. Its area under the receiver operating characteristic curve (AUC-ROC) is 0.737 against 0.670 for the best transformer on the primary seed, and no added text score detectably raises the baseline's AUC-ROC. None of the five domain-pretrained transformers detectably beats its general-domain control after Bonferroni correction. For long-horizon prediction, the baseline remains the model to beat.
Read the full abstract

All authors Daniel Kua, Emrul Hasan, John-Jose Nunez, Frances Chen

Artificial Intelligence

ST-Bench: A Spatial-Temporal Benchmark for Multi-Agent System Generation on Scientific Research Tasks

Qi Cheng, Rongchao Dong, Shengyu Chen et al. (+8 more)

The rapid progress of LLM-based multi-agent systems (MAS) has shown that they largely outperform single agents on coding, math, and QA tasks, where executable tests provide a binary success signal. Whether this advantage transfers to real scientific data analysis remains untested. We introduce ST-Bench, a benchmark designed to answer two questions: whether MAS outperform single agents on complex scientific data analysis tasks, and if so, by how much and at what additional cost. ST-Bench contains 100 data science tasks adapted from published Earth science studies across hydrology, agriculture, and wetland methane research, expanded into 2,067 queries grounded in additional published studies and validated by domain experts. Using ST-Bench, we evaluate five recent MAS generation methods under two training protocols, against single-agent baselines on the same GPT-5 backbone. Nine of the ten MAS configurations exceed the cheapest single-agent baseline, with the strongest reaching nearly three times its composite score. This gain is primarily attributable to coverage: trained workflows produce realistic numerical metrics on a larger fraction of queries, while the quality of those metrics, conditional on producing realistic output, is comparable to that of the single-agent baseline. The strongest configuration requires approximately four times the single-agent inference time, whereas a more economical workflow captures the majority of the benefit at less than twice the cost. MAS specialization confers measurable benefit on scientific data analysis, but the benefit is conditional rather than universal.
Read the full abstract

All authors Qi Cheng, Rongchao Dong, Shengyu Chen, Licheng Liu, Dan Lu, Zhengzhang Chen, Wei Cheng, Yiqun Xie, Haifeng Chen, Xiaowei Jia, Haoyu Wang

Software Engineering

ES-Trace: Auditing Ethical-Sourcing Disclosure of Code Generation Models Beyond Model Cards

Zhuolin Xu, Haibo Wang and Shin Hwei Tan

Code generation models have been increasingly used in software development, but their development raises ethical-sourcing concerns involving intellectual property, privacy, fairness, labour practices, and environmental impact. Although prior work has defined ethical-sourcing criteria for code generation, it remains unclear how much evidence existing models disclose and where that evidence can be found. We introduce ES-Trace, a framework for ethical-sourcing disclosure audits that traces disclosed evidence beyond model cards using the Model Documentation Traceability Graph (MDTG), which represents relationships among models, versions, and documentation artifacts. We apply ES-Trace to 26 models from 10 publishers across 77 documents and 20 ES-CodeGen aspects. Model-card-only auditing yields a mean score of 1.77/5, while resolving the declared references increases it to 2.82/5, with most of the increase arising from documents that the publisher declares in structured metadata. The key findings of our study include: (1) social and labour-related aspects remain poorly documented, even when expanding the audit to the full documentation scope, (2) resolving documentation references substantially increases observed disclosure, raising the mean score from 1.77/5 to 2.82/5, (3) model-card-only audits can mischaracterize release-level disclosure changes, and (4) documentation mismatches can associate evidence with the wrong model or version, highlighting the need for explicit model–version binding and consistency across documentation artifacts. Our study calls for reference-aware ethical-sourcing disclosure audits, explicit model–version binding and consistency across documentation artifacts, and stronger documentation of currently underreported social and labour-related aspects.
Read the full abstract

Information Retrieval

Contrastive Learning for Aspect Representation towards Explainable Recommendation

Emrul Hasan and Chen Ding

In this work, we propose a novel recommendation model, CLARER (Contrastive Learning for Aspect Representation towards Explainable Recommendation) that integrates aspect features learned from textual reviews with rating information to improve the accuracy and explainability of recommendations. Our proposed framework learns user and item representations by combining rating-based features and aspect-based features from reviews. Specifically, rating-based features are learned through a multi-layer perceptron (MLP) model, while aspect-specific review representations are learned using a transformer encoder to capture the semantic information and contrastive learning to better distinguish user preferences. To provide explanations, we train a transformer decoder, using the final representations of users and items from both rating and aspect-based features as context. Experimental results in three benchmark data sets demonstrate that our model achieves superior performance compared to baseline methods in both recommendation (accuracy) and explanation generation.
Read the full abstract

Information Retrieval

Token-Budgeted Escalation for Financial Document Question Answering: Cost Is Predictable, Benefit Is the Bottleneck

Junru Zhu, Yixin Yang, Xiaoqing Ding et al. (+1 more)

Retrieval-augmented generation systems can route difficult queries to deeper context, but batch deployments must allocate a shared token budget across calls whose costs vary by query. We formulate selective escalation as finite-batch allocation for financial document question answering. Each of 150 FinanceBench questions first receives a top-1 retrieval answer. Predictors estimate the adjudication-quality gain and token cost of an optional top-5 call, and the allocator prioritizes calls by predicted gain per token. At the nominal 10% budget, gain-per-token allocation improves adjudication quality over gain-only ranking by 0.034 (95% document-bootstrap confidence interval [0.001, 0.072]) while using 46.6% fewer total tokens than one-pass top-5 retrieval. Additional-call cost is accurately predictable (R-squared 0.93), whereas beneficial escalation remains difficult to rank (area under the receiver operating characteristic curve 0.60). These results show that heterogeneous cost is actionable under tight constraints, while progress across the full budget frontier depends on stronger query-specific benefit estimates.
Read the full abstract

All authors Junru Zhu, Yixin Yang, Xiaoqing Ding, Ruoyu Qi

Distributed Computing

When Can Stateless Recovery Defeat Byzantine Quorum Safety? A Tight Normal Form for Single-Step BFT

Arnab Mallick and Indraveni Chebolu

Byzantine quorum safety relies on correct replicas refusing to sign conflicting values. A replica that loses its protocol state during recovery but retains its identity and signing key may forget an earlier vote. We study certificates formed by matching signed votes from at least $q$ of $n$ replicas, assuming that each correct replica avoids conflicting votes between recoveries. If two conflicting certificates form, their overlap has size between $2q-n$ and $b+c$, where $b$ counts Byzantine replicas and $c$ counts correct identities that recovered during the execution considered. Our main result decomposes the slack $b+c-(2q-n)$ into four nonnegative counts: extra signers in the first certificate, extra signers in the second, identities in neither certificate, and Byzantine or recovering identities outside their overlap. Zero slack forces an exact signer partition. With $n=3f+1$ replicas, threshold $q=2f+1$, at most $f$ Byzantine replicas, and exactly one correct recovery event, any conflicting pair forces exactly $f$ Byzantine replicas, all in the overlap together with the recovered replica; each certificate has a disjoint side of $f$ correct replicas. A minimal protocol attains this form. We distinguish certificate formation from acceptance, give a sufficient check using configured fault and recovery caps, and explain why durable vote records written before signature release prevent the conflict.
Read the full abstract

Computer Vision

Later Is Better: Token Reduction for ViTs Under Distribution Shift

Hyeongheon Cha, Hyungjun Yoon and Sung-Ju Lee

Training-free token reduction accelerates vision transformers by removing redundant tokens across layers, recovering most of the original accuracy at a fraction of the compute. These methods, however, are designed and evaluated primarily on clean data, and under real-world distribution shift their accuracy gap to the uncompressed model widens with the removal rate. We show that this gap is governed by the reduction schedule, the depth profile of removal, usually left fixed as an implementation detail. Concretely, we introduce a one-parameter late-concentrated power-law schedule that consistently improves out-of-distribution accuracy over flat at no extra inference cost. On ImageNet-C with DeiT-S, the late schedule closes 83% of that gap at a 26% compute reduction (+1.17pp), and 99% of it at a lighter 7% reduction (+0.26pp). The gain cannot be attributed to retaining more tokens or using extra compute: held to flat's compute, the late schedule removes more tokens in total and leaves fewer tokens at the end, yet still wins. Single-layer probes point to a mechanism: earlier reductions perturb features that pass through more remaining layers, front-loading reduction error in depth. The effect is broad, holding across five token-reduction methods (ToMe, EViT, ATS, ATC, PiToMe), nine backbones, all ImageNet-C corruption types, eight further shift suites, and two further modalities, video and vision-language QA. It is also specific to shift, still positive on clean and rising monotonically to ~4x that at the highest severity 5. The schedule keeps its gain under six test-time adaptation methods, and needs no per-input or per-domain tuning.
Read the full abstract

Software Engineering

Acquiring and Verifying Repository Norms for Coding Agents

Kaifeng He, Xiaojun Zhang, Zhenxi Chen et al. (+2 more)

Changes produced by coding agents can pass functional tests while leaving repository contribution requirements unmet. Following repository-specific norms requires identifying guidance dispersed across repository sources and interpreting its conditions and exceptions. Retrieval and documentation approaches supply general context, but agents must still determine which norms apply. We introduce RepoNorm to acquire explicit and implicit repository norms independently of coding tasks. It checks norm content and applicability using repository evidence, consults Git history when needed, and delivers norm packages to existing coding agents. Our evaluation uses three coding models and 121 tasks from RepoNormBench. Against the baseline with no additional generated guidance (Raw), relative improvements are 7.42-10.77% for Overall Norm Compliance Rate (NCR), 31.64-45.44% for Contribution NCR, and 11.34-17.69% for Prompt-omitted NCR. All three coding models also obtain higher values on these NCR measures with RepoNorm than with CodeWiki documentation. Functional success rates show observed gains of 5.79-9.09 percentage points over Raw; the paired comparisons do not reach the significance threshold. Sampled precision is 86% with RepoNorm's default configuration.
Read the full abstract

All authors Kaifeng He, Xiaojun Zhang, Zhenxi Chen, Christoph Treude, Mingwei Liu

Robotics

StairVLA: Stage-Aware Hierarchical Action Generation for Vision-Language-Action Models

Shangyuan Yuan, Xinda Qi, Yujiang Pu et al. (+2 more)

Vision-language-action (VLA) models increasingly rely on diffusion- or flow-matching-based action heads to generate continuous robot actions. These action heads typically process the denoising trajectory in a largely uniform manner. However, we observe that the conditioning focus naturally shifts across denoising stages: early stages combine language instructions and visual observations to establish a coarse action trajectory, whereas later stages place greater emphasis on current visual observations for action alignment. Based on this insight, we introduce StairVLA, a stage-aware hierarchical action generation framework that uses partially denoised actions as a natural interface between coarse long-horizon action generation and local refinement. A high-level VLA performs early denoising to produce a reusable long-horizon partially denoised action trajectory, while a lightweight refiner operates at a higher frequency to refine local action chunks using the latest observations. This design amortizes expensive high-level VLA computation while preserving frequent closed-loop correction. On LIBERO, our GR00T-style instantiation improves average success from 96.5% to 97.8% while reducing amortized inference latency from 115.0 ms to 44.2 ms per action chunk. More broadly, across two VLA backbones, simulation benchmarks, and real-robot tasks, StairVLA consistently reduces inference cost while maintaining strong task performance.
Read the full abstract

All authors Shangyuan Yuan, Xinda Qi, Yujiang Pu, Wenliang Guo, Xiaobo Tan

Machine Learning (Statistics)

Trustworthy Method Comparison with AI Judges: Estimation and Design under Order, Batch, and Aggregation Effects

Tianxi Li and Jie Ding

Large language models (LLMs) are increasingly used as judges for automated AI evaluation. A common practice is to randomize prompt sequences and average the resulting scores, but its statistical validity remains unclear. We show that LLM evaluation mechanisms can be approximated by a class of Markov generalized linear mixed models (GLMMs), supported by out-of-sample predictions across three major commercial LLMs. Using a first-order Markov GLMM, we study leaderboard ranking and group comparison. For leaderboard ranking, randomize-and-average selection is consistent under a mild separation condition, and a Williams square design can improve efficiency when item qualities are close. For group comparison, naive averaging can yield inconsistent conclusions about differences in group-level quality because of the response model's nonlinearity. Empirical results further support the validity of the proposed model-based inference beyond the first-order theory, including settings with higher-order sequence memory. We illustrate the approach in an application where AI judges compare two graphical model estimation methods.
Read the full abstract

Machine Learning

Adversarially Trained Linear Transformers Are Optimal Robust In-Context Learners for Gaussian Mixtures

Soichiro Kumano

Adversarial training is one of the most reliable defenses against adversarial attacks, but its high computational cost must generally be paid anew for each task. Robust foundation models offer a promising alternative: adversarially pretrain a model once and then transfer its robustness to downstream tasks through lightweight adaptation. However, a fundamental question remains open: can robustness acquired during pretraining transfer to unseen tasks without further adversarial training? In this study, we answer this question affirmatively. A single model adversarially pretrained at scale can achieve optimal robustness on new tasks without additional task-specific training. Specifically, we show that, for a family of Gaussian-mixture classification tasks, a sufficiently deep linear transformer adversarially trained across tasks can asymptotically attain the robust Bayes error on previously unseen tasks through in-context learning from clean demonstrations. By contrast, a standardly trained model cannot. We further analyze convergence under gradient flow, an accuracy–robustness trade-off, and demonstration complexity.
Read the full abstract

Computation and Language

From Evidence to Action: How Tool-Using Agents Fail

Hongzhan Lin, Shidong Cao, Ziyang Luo et al. (+3 more)

Tool-using agents make consequential changes to external state, yet correct outcomes do not guarantee that their actions were supported by evidence established beforehand. We study where this evidence-to-action chain breaks as agents move from deciding whether to act to executing single actions and dependent workflows. Across ten model-harness configurations, strong static action assessment can coexist with much weaker interactive execution. Failures often begin before execution: agents stop with incomplete investigation or act before required evidence is established. Once required evidence is obtained, single-action execution is usually reliable, while multi-action workflows additionally expose unresolved prerequisites and incomplete execution. For this analysis, we introduce SafeActBench, comprising 656 cases across six operational domains and five protocols that progress from static action judgment and investigated non-action to single- and multi-action workflows. A provenance-bound Evidence Ledger and deterministic trajectory evaluator track what information was established, when actions occurred, and whether downstream dependencies were satisfied. These results show that failures arise not only from missing information, but also from how agents use established evidence when deciding and executing actions.
Read the full abstract

All authors Hongzhan Lin, Shidong Cao, Ziyang Luo, Wenhao Chai, Mong-Li Lee, Wynne Hsu

Robotics

CoRE: Learning Collaboration-Role Experts for Decentralized Collaborative Manipulation with One Policy

Yanan Zhou, Zhaoyan Qian, Zihao Li et al. (+3 more)

Collaborative manipulation requires robots to perform complementary actions as interactions unfold. We study single-policy decentralized collaboration: every robot runs the same policy from its visual observations and proprioception, without task prompts, identity labels, or inter-robot messages. The challenge is to learn complementary team behaviors within shared parameters and select appropriate actions from each robot's local observations. We introduce CoRE, which learns Collaboration-Role Experts from pooled multi-task, multi-robot demonstrations. Fused appearance and geometry provide local interaction evidence. Query-conditioned cross-attention experts provide adaptable prediction paths, which a local router combines at each action-chunk position. During training, an action-expert alignment loss supervises expert selection using relative forced-route prediction errors against demonstrations under fixed inputs, without role labels. Across simulation benchmarks, CoRE achieves the highest average performance among evaluated decentralized methods. Physical experiments demonstrate effective collaboration across diverse manipulation tasks and robustness to partner delays and slowdowns. Project page: https://aus.bot/research/core/.
Read the full abstract

All authors Yanan Zhou, Zhaoyan Qian, Zihao Li, Mingyuan Ba, Ranpeng Qiu, Weiming Zhi

Artificial Intelligence

How Well Do LLMs Reason with Noisy Evidence? An Active Visual Reasoning Benchmark

Bach Nguyen, Zhaonan Li, Mau Son Nguyen et al. (+5 more)

Real-world reasoning rarely reduces to static question answering: agents must actively gather information from tools and sensors that are often noisy and unreliable. Yet most existing active reasoning benchmarks assume that environmental feedback is trustworthy, or introduce noise without exposing an explicit, calibrated uncertainty signal, leaving open how LLMs should reason when the evidence itself is uncertain. We introduce VisualNoiseQA, a novel benchmark for active reasoning under noisy visual feedback. A text-only LLM must solve VQA problems by iteratively querying a fixed, off-the-shelf VLM treated as a stochastic visual sensor. For each query, we draw multiple samples and expose an empirical uncertainty signal via self-consistency, enabling the reasoner to probe from different angles and decide what to ask next and when to stop. Our construction is automatic and scalable: starting from diverse VQA sources and two noisy VLMs, we retain only questions where the sensor is inconsistent yet human-solvable. We evaluate multiple LLM reasoners on 1,000 instances spanning perception, chart understanding, and knowledge-intensive reasoning. VisualNoiseQA thus provides a controlled playground to study how different LLMs exploit uncertainty signals for robust reasoning.
Read the full abstract

All authors Bach Nguyen, Zhaonan Li, Mau Son Nguyen, Sanika Chavan, Nilay Kumar, Hong Anh Nguyen, Khoa Vo, Ben Zhou

Optimization and Control

PLADOS: A Parameter-free Landing Algorithm for Decentralized Optimization on the Stiefel Manifold

Shu Li and Jiang Hu

Decentralized optimization on the Stiefel manifold has broad applications in machine learning and signal processing. We propose PLADOS, a parameter-free landing algorithm for decentralized optimization on the Stiefel manifold. Each iteration combines local stochastic gradient updates with a single retraction-free communication step over the communication graph. A key convex-like property of the intersection of the nonconvex Stiefel manifold and consensus constraints is characterized through a restricted secant inequality for the penalty-only scheme, ensuring local contraction of the manifold consensus error without introducing additional parameters that require tuning. For $n$ nodes using independent local batches of size $b$, we prove that PLADOS achieves linear speedup with respect to the number of nodes with an asymptotic convergence rate of $O(\sigma/\sqrt{nbK}+(n\xi^2)^{1/3}/K^{2/3})$ after $K$ iterations, where $\sigma^2>0$ is the sampling variance and $\xi^2$ quantifies gradient heterogeneity. Experiments demonstrate stability across tested stepsizes and networks, time efficiency, and linear speedup with respect to the number of nodes.
Read the full abstract

Databases

Cost-Effective Numerical QA over Semi-Structured Table: Structuring, Resolution, Planning

Feng Luo, Hui Luo, Zhifeng Bao et al. (+3 more)

Semi-structured tables encode rich semantic information through diverse layout elements, such as hierarchical row and column headers. Answering numerical questions over such data is challenging because it requires a joint effort of accurate structural understanding of tables, question uncertainty resolution, and complex query intents interpretation. Existing methods are rarely effective in handling the above multifaceted challenges in a holistic manner; moreover, they rely on LLMs without considering financial implications. We propose SemiBonsai, a cost-effective, LLM-powered framework featuring: (1) a Multiway Layered Table Structurer that converts a table into a multiway layered tree organized by structural elements; (2) a Context-Aware Uncertainty Resolver that grounds underspecified phrases to context-coherent structural elements; (3) a Plan-Guided Reasoner that decomposes complex query intents into a query plan for reasoning; (4) a Budget-Aware Bandit Router that learns per-instance LLM utilities and optimizes LLM selection under fixed monetary constraints. Experiments on three benchmarks show that SemiBonsai improves both QA effectiveness and cost-effectiveness. Notably, the Table Structurer improves answer accuracy by 46% over alternative tree construction methods, and Router achieves a 5% average relative gain in answer accuracy compared with existing routing methods across budgets.
Read the full abstract

All authors Feng Luo, Hui Luo, Zhifeng Bao, J. Shane Culpepper, Xiaoli Wang, Shazia Sadiq

General Relativity and Quantum Cosmology

Charged soliton–black hole phase transitions in three-dimensional Einstein–Gauss–Bonnet gravity

Cendikiawan Suryaatmadja and Robert Mann

We construct a magnetic soliton by double Wick rotating the static electrically charged black hole in three-dimensional Einstein–Gauss–Bonnet gravity. We match the boundary metrics and Maxwell sources under the chosen boundary condition and compute the difference of the renormalized Euclidean actions. Because the Maxwell potential grows logarithmically at infinity, the chosen finite Maxwell boundary term changes both the fixed Maxwell source and the renormalized action. The charge contributions for both the soliton and the black hole appear with plus signs in $\Delta I^{(c_{\mathrm{fin}})}$. For the grand-canonical boundary condition $c_{\mathrm{fin}}=1$, the black-hole and soliton branches can cross with different slopes. For the example considered below, the black-hole branch satisfying the response conditions ends at a maximum temperature, and the coexistence curve has a $T\to0^+$ endpoint at two opposite values of the spatial Maxwell source. At a nonextremal crossing, the entropy difference is $S_b-S_s=S_b$. For these two static solutions, equality of both current components requires both charges to vanish.
Read the full abstract

Networking

When Weak Reports Matter: Staged Anchored Fusion for Cooperative UAV Sensing

Xingqin Lin

Local multipath rejection can erase evidence needed for cooperative sensing. We propose staged anchored recovery: preserve strong-only confirmations, then query compatible weak reports using unused strong anchors. For any number of sensing nodes, we prove lossless residual screening and derive corroboration and bidirectional cost laws. In 1,024 five-UAV drops, recovery adds 30 matched targets and three false outputs over strict consensus, matching one-pass anchored confirmation's detection counts while reducing weak uploads by 97.4%. Equal-sized cue/report records yield 4.5% less payload than uploading all eligible reports. Independent validation recovers four additional targets with no observed false outputs.
Read the full abstract

High Energy Physics (Theory)

Thermodynamics and mechanical laws of causal-diamond horizons

Carlos R. Ordóñez and Gustavo Valdivia-Mera

We develop a unified thermodynamic and mechanical framework for causal-diamond horizons in Minkowski spacetime. Euclidean regularity fixes the temperature $T_D=1/(\pi\alpha)$, where $2\alpha$ is the temporal extent of the diamond. The regulated Gibbons-Hawking-York contribution gives $I_D^E=-A_0/(4G)$, with $A_0$ the area of the horizon bifurcation surface, and the vanishing equilibrium geometric energy then yields the entropy-area relation $S_D=A_0/(4G)$. For weak matter perturbations crossing a localized portion of the future horizon, a Raychaudhuri analysis relates the associated Killing-energy flux to the area response, yielding a local physical-process first law at leading order near the bifurcation surface and, for classical matter satisfying the null energy condition, a local second law. Uniform conformal surface gravity establishes the zeroth law, while zero temperature is approached only in the limit $\alpha\to\infty$, in which the finite causal horizon disappears. These results show that a coherent horizon-law structure can arise within a causal diamond, a fundamental building block of the causal structure of Minkowski spacetime. The derivation is explicit in $(2+1)$ dimensions, while the physical results retain the same form for spherical causal diamonds in $d\geq3$.
Read the full abstract

Robotics

Seeing Through the Displaced Frame: Privileged Noise Distillation for Vision-Force Precision Assembly

Ching-Hsiang Chang, Tzu-Yu Chuang, Yi-Hsiu Lee et al. (+3 more)

Pose error in precision assembly can corrupt not only what a robot observes but also the coordinate frame in which it acts. On the FORGE benchmark, the official state-based policy succeeds in 97% to 99% of episodes with the true pose but only 32% to 60% at the benchmark's $\sigma=5$ mm pose-noise setting. The same estimated pose enters the observation and anchors the action frame, making the offset unidentifiable from proprioceptive state alone before contact. We supply this missing information during training in two ways. A privileged teacher observes the offset in simulation, while clean demonstrations can instead be relabelled into the displaced frame in closed form. The deployed student is trained with behaviour cloning followed by one DAgger round and receives only noisy state, a raw wrench window, and two RGB cameras at test time. On the unmodified FORGE tasks, the teacher-route student maintains 92% to 99% success across $\sigma=0$ to 5 mm, while six non-privileged baselines fall to 2% to 80%. A matched behaviour-cloning experiment isolates the source of this robustness. With the same student architecture, data budget, and training procedure, demonstrations generated without offset access yield only 31.5% success at $\sigma=5$ mm, whereas privileged and relabelled demonstrations reach 88.7% and 95.8%. Deployed zero-shot on a Franka, the student reaches 83.3% pooled success at $\sigma=5$ mm against 34.4% for the strongest state-based policy. Deployable sensing alone is insufficient. Robustness requires supervision that encodes compensation for the latent frame offset.Project page: https://drychang.github.io/displaced-frame/
Read the full abstract

All authors Ching-Hsiang Chang, Tzu-Yu Chuang, Yi-Hsiu Lee, Yi-Ting Chen, Yuan-Fu Yang, Min Sun

Human-Computer Interaction

A Pedagogically Demonstrative Model Visualizing the Pathway from Online Interactions to Personalized Recommendation

Sushmita Khan, Connor Pennington and Bart P Knijnenburg

Personal digital activity increasingly shapes online experiences, yet few users have been educated regarding the processes transforming raw interactions into personalized suggestions. We developed an education artifact that illustratively simulates how AI leverages users' digital activities to shape online recommendations (e.g., ads). Our artifact processes users' digital activity using a locally-hosted LLM to generate user profiles of their inferred interests and personalized recommendations. A three-layered Sankey diagram maps data sources through inferred interests to personalized recommendations. Interactive filters enable users to explore how different combinations of data sources influence personalized outcomes. This paper describes the artifact and its educational value, and reports findings of a pilot think-aloud study with six young adults. We find that the artifact effectively taught participants the conceptual relationship between digital activities and personalized recommendations. While this did lead participants to develop privacy awareness, they anticipated minimal behavior change due to the perceived unavoidability of platform participation.
Read the full abstract

Instrumentation and Methods for Astrophysics

Very Long Baseline Interferometry with the SKA Observatory

J. F. Radcliffe, J. Yang, Z. Paragi et al. (+131 more)

Very Long Baseline Interferometry (VLBI) is a fundamental technique that combines signals from telescopes separated by thousands of kilometres to obtain some of the highest-resolution images in astronomy. The Square Kilometre Array Observatory (SKAO) will provide sensitive elements to current VLBI arrays, enabling deep VLBI observations (including the first low-frequency VLBI capability in the Southern Hemisphere). This summary chapter outlines the science that the SKAO, when participating in a VLBI array, will achieve (such as AGN physics, explosive transient localisation, Galactic structure, and gravitational waves), and the current understanding of how VLBI with the SKAO will operate. Given that VLBI requires partnerships with other instruments, the chapter concludes with outstanding issues and considerations for the community needed to maximise the scientific return on investment in SKA-VLBI.
Read the full abstract

All authors J. F. Radcliffe, J. Yang, Z. Paragi, O. Bayandina, C. Reynolds, C. García-Miró, T. An, M. Abada-Simon, J. N. H. S. Aditya, I. Agudo, T. Akahori, A. Alberdi, S. Anton, P. Atri, A.-K. Baczko, S. Belladitta, E. K. Bempong-Manful, R. Beswick, M. Bhardwaj, E. Bonnassieux, A. Brunthaler, C.-S. Chang, P. Charlot, S. Chen, W. Chen, X. Cheng, A. Chhipa, J. O. Chibueze, I. Cho, J. B. Climent, L. Cui, S. Curiel, F. D'Ammando, J. R. Dawson, A. Deller, H. Ding, R. Dodson, J. Drevon, F. Eppel, S. Frey, M. Garrett, M. Gawroński, J. Geng, S. Giarratana, M. Giroletti, S. Gulyaev, L. I. Gurvits, C. Hao, T. Hirota, A. J. Hedge, H. Hu, H. Imai, P. Jiang, J. de Jong, M. Kadler, B. Karimi, P. Kharb, R. Khatun, J.-Y. Kim, J. Kim, J. Kim, F. Kirsten, H. R. Klöckner, H. Kobayashi, Y. Y. Kovalev, M. Kunert-Bajraszewska, A. Kutkin, H. J. van Langevelde, J. Li, Y. Li, Y. Li, Z. Lin, D. Liu, H. Liu, K. Liu, A. Lobanov, R.-S. Lu, V. Mahatma, B. Marcote, M. Mezcua, J. Moldón, L. K. Morabito, S. E. Motta, S. Murthy, I. Natarajan, A. Njeri, J. Oh, M. Orienti, A. Osetrova, M. Pandey-Pommier, F. Panessa, J. Park, M. Pérez-Torres, L. Petrov, J. C. S. Pierce, C. Phillips, L. Ricci, M. Rioja, H. Ro, J. Röder, E. Ros, M. Rupen, K. L. J. Rygl, R. Santra, S. Silpa, T. Savolainen, G. A. S. Sundaram, P. Shastri, X. Shu, A. Słowikowska, B. W. Sohn, C. Spingola, F. Sweijen, K. Takahashi, R. Timmerman, F. Ubertosi, L. Uscanga, E. Varenius, M. Verkouter, T. Vernstrom, S. Wagner, D. Williams-Baldwin, J. Wongphechauxsorn, X. Wu, M. Xu, Y. Xu, M. K. Y{\i}ld{\i}z, Z. Yan, A. Yang, X. Yang, S. Yao, W. Yu, B. Zhang, P. van Zyl

Programming Languages

Cleave: Scaling Tensor Program Optimization via Decoupled Algebraic Search and Operator Scheduling

David Pissarra, Jinkun Lin, Haitian Jiang et al. (+2 more)

Optimized kernels such as FlashAttention and FlashDecoding are crucial for accelerating today's large models. Most of them are handwritten by experts because existing ML compilers cannot match their efficiency. Producing such kernels requires fusing computations with multiple reductions, which requires both algebraic transformation of the computation graph and operator scheduling of the transformed graph. Unfortunately, searching the two jointly yields a space too large to navigate. We propose Cleave, an ML compiler built on symbolic decoupling: Cleave discovers transformations by performing superoptimization on a graph with symbolic shapes, and then schedules each resulting graph on concrete shapes. Representing shapes as symbols makes equivalence checking cheap and lets a new Split operator, with a symbolic split count, parallelize along a reduction dimension. Cleave's scheduler fuses graphs with multiple reductions through iterative tiling and horizontal fusion. Evaluation on common LLM subgraphs shows that Cleave generates kernels up to 2.8x faster than the best baseline (1.6x on average) and reduces compilation time by 5.9x on average compared to Mirage. For dynamic workloads captured from production serving traces, Cleave compiles each operator once and achieves geometric mean speedups of 1.4x and 1.7x over FlashInfer's handwritten FA2 and FA3 backends. Cleave's code is available at: https://github.com/nyu-systems/cleave
Read the full abstract

All authors David Pissarra, Jinkun Lin, Haitian Jiang, Aurojit Panda, Jinyang Li

Probability

Sharp Asymptotics for the Solvability Probability of Random Stable Roommates

Caden Young and Ian Studebaker

For even $n$, let $P_n$ be the probability that independent uniform strict preference lists on $n$ participants admit a stable perfect matching. We prove \[ P_n\sim\frac{e\,2^{1/4}\Gamma(3/4)}{\sqrt\pi}\,n^{-1/4}. \] This establishes Mertens's conjectured exponent of decay, with a leading constant different from his original numerical prediction. The proof starts from Mertens's exact alternating sum over stable permutations. To preserve its cancellation, we construct a common approximation for all cycle structures with the same number of vertices in cycles longer than two. By symmetry, the integrated first-order correction is the same for every such cycle structure, and the remaining errors can be summed in absolute value. The enumeration then reduces the probability to a one-dimensional sum with positive terms.
Read the full abstract

Machine Learning (Statistics)

High-dimensional online calibration from harmonic weights

Maxwell Fishelson and Mehryar Mohri

We study the online calibration of multidimensional forecasts over an arbitrary convex set $Y\subseteq\mathbb{R}^d$ relative to an arbitrary error norm $\|\cdot\|_{L}$. For forecasting $d$ binary outcomes simultaneously ($Y=[0,1]^d$), we give the first algorithm that achieves $\varepsilon$-calibration in a number of rounds that is polynomial in $d$ for every fixed accuracy. It requires $d^{O(1/\varepsilon)}$ rounds, exponentially improving the dimension dependence of previous bounds. For multi-class forecasting ($Y=\Delta_d$), we obtain the same $d^{O(1/\varepsilon)}$ rate, improving the $d^{\widetilde{O}(1/\varepsilon^2)}$ bounds of Peng and Fishelson et al. Our algorithm is simple: on each round, it outputs a harmonically weighted distribution over harmonically smoothed past outcomes. The same algorithm works for every forecast set and norm. More generally, it achieves $\varepsilon$-calibration after $\exp(O(\gamma(Y,L)/\varepsilon))$ rounds, where $\gamma(Y,L)$ is a geometric parameter defined by a matrix discrepancy problem. The harmonic weights are motivated by the fact that the discrete Hilbert transform matrix achieves the optimal discrepancy up to a universal constant, simultaneously for every $L$. This optimality result may be of independent interest.
Read the full abstract

Machine Learning

Cite What You Explore: Budget-Aware LLM Reasoning over Medical KGs with Verifiable Evidence

Chen Chen, Dongjie Wang, Mei Liu et al. (+1 more)

Post-discharge risk prediction from electronic health records (EHRs) is difficult because many dependencies that link discharge-time observations to downstream complications, such as comorbidity cascades and drug-disease interactions, are absent from the record. External medical knowledge graphs (KGs) can supply these missing dependencies, but tracing them demands three properties: KG exploration must remain cost-bounded, retrieved evidence must be differentiated by source quality, and the resulting rationale must be citable for retrospective review. Large language models (LLMs) can plan and verify over structured evidence, making them natural candidates for KG reasoning, but existing LLM-based methods do not satisfy these three properties jointly. In this paper, we propose BAR, a Budget-Aware LLM Reasoning framework over medical KGs with three contributions. First, BAR refines the raw KG into disease-specific evidence graphs whose edges carry support scores and provenance records, turning the KG into a quality-annotated reasoning space rather than a static feature source. Second, an LLM then reasons over this graph through a plan-navigate-verify loop that decomposes the question into steps, retrieves evidence under a patient-specific budget, and revises when verification fails. Third, a reasoning policy is trained with a reward that compares predictions with and without acquired evidence, combined with acquisition cost and citation-integrity terms. Across 8 diseases and 3 prediction horizons on MIMIC-III and MIMIC-IV, BAR improves AUPRC by 3.4 points over the strongest baseline, raises citation precision from 59.8% to 77.9%, and consumes only 62-65% of the budget cap.
Read the full abstract

All authors Chen Chen, Dongjie Wang, Mei Liu, Zijun Yao

Analysis of PDEs

Local Propagation of Momentum Moments for the Relativistic Vlasov–Poisson System

Zhaopeng Wang

This paper proves local propagation of momentum moments of order $k\ge5/2$ for the three-dimensional relativistic Vlasov–Poisson system. In the Coulomb case, this improves the previous result which requires integer moments of order $k\ge3$. The proof adapts the critical density method developed in our recent work on the classical Vlasov–Poisson equation. We introduce a weighted kinetic tensor adapted to the relativistic velocity and obtain the critical density estimate. This estimate then allows us to close the moment argument by a short-time bootstrap.
Read the full abstract