Robotics
SkillX: Unified Multi-Skill Policy Learning for Humanoid Soccer
Zhangchen Ye, Enxuan Ruan, Yifei Bao et al. (+10 more)
Humanoid soccer is a challenging testbed for dynamic whole-body control, requiring
robots to coordinate balance, locomotion, object interaction, and skill switching over long horizons. Existing
humanoid sports methods often rely on task-specific multi-stage pipelines, making it difficult to jointly learn and compose multiple object-interactive skills within a single deployable policy. To address this, we present SkillX, a unified reinforcement learning framework that learns and composes multiple atomic soccer skills through a single command-conditioned policy. SkillX integrates three core designs: skill-specific
adversarial motion priors, skill-specific critics, and an object-aware temporal encoder, enabling the
robot to execute atomic skills and transition among them such as dribbling, trapping, and shooting. Experiments in simulation and on a real Noetix E1
humanoid demonstrate robust multi-skill execution, long-horizon skill composition, and successful sim-to-real deployment.
Read the full abstract ▾
All authors Zhangchen Ye, Enxuan Ruan, Yifei Bao, Runhan Huang, Jiankun Yang, Jiakang Jin, Yixiao Huo, Pengyuan Wang, Yinan Han, Huaxing Huang, Wenhao Cui, Yiming Li, Xiaoyu Tian
Systems and Control
Moment Matching for Descriptor Systems: A Möbius Mapping Approach
David Militaru, Denisa Jungheatu, Andrei Sperilă et al. (+1 more)
For a class of single-input single-output systems described by proper or improper transfer functions, we propose moment-matching procedures applicable in both continuous- and discrete-time contexts. The resulting technique is not only more flexible and reliable than other procedures which are currently available in literature, but it also enables the placement of constraints on the reduced-order model's poles and zeros. This constraint-based feature, hitherto available only for continuous-time state-space systems, is illustrated via a numerical example based on a practical problem from literature.
Read the full abstract ▾
All authors David Militaru, Denisa Jungheatu, Andrei Sperilă, Tudor C. Ionescu
Artificial Intelligence
We Built a Mirror and Mistook It for a Mind: Causal Liability and the Fallacy of AI Consciousness
Afshin Khadangi
The contemporary debate over machine consciousness begins from a concealed assumption: that the object called "AI" already constitutes the kind of entity to which consciousness could belong. This paper challenges that assumption by separating phenomenal consciousness, introspective report, and human projective introspection, then arguing that generative systems can return linguistic traces of human interiority in first-person form without thereby identifying a phenomenal bearer. We call the resulting inference the AI Consciousness Fallacy. We then introduce Causal Liability Theory (CLT). CLT-I proposes liability closure as a criterion for individuating a candidate bearer: a physically continuing process becomes the non-delegable inheritor of constraints generated by its own endogenous discriminations. CLT-II advances the stronger
conjecture that liability closure is necessary and sufficient for minimal phenomenal subjecthood. An open-weight causal audit operationalizes CLT-I across multiple model families. Forced discriminations produced persistent downstream divergence; activation patching showed strong causal mediation; live and copied adaptive states were behaviorally identical under matched randomness; and detached reconstruction preserved computational state across process replacement while, by protocol, breaking constitutive continuity and non-delegable inheritance. These results show that CLT-I distinctions are experimentally tractable and can dissociate causal bearer structure from first-person performance. The framework therefore separates consciousness attribution, causal bearer individuation, and the independent metaphysical question of consciousness constitution.
Read the full abstract ▾
Systems and Control
Local Energy Function-Based Power System Transient Stability Analysis with Grid Forming and Grid Following Inverters
Yingyi Tang, Junbo Zhao and Baosen Zhang
The widespread deployment of inverter-based resources (IBRs) is reshaping power system dynamics by their control strategies. Grid-following (GFL) and grid-forming (GFM) inverters are expected to co-exist from now on. However, ensuring the transient stability of such systems remains a critical and underexplored challenge. In this paper, a comprehensive transient stability analysis of the mixed-GFM-GFL inverter system is conducted using a newly derived local energy function. The transient stability analysis model of the system is first derived to reveal the dynamic interactions between GFL and GFM inverters during transient responses. A local energy function is then proposed for analytical transient stability assessment. Instead of ignoring or making harsh assumptions on inverters' coupled damping effect, the proposed local energy function carefully addresses its impact on system transient stability by locally bounding the region where the system's energy dissipation property holds. Such a property fully considers the combined damping effect of GFL and GFM inverters. The derived critical energy and stability index provide a fast and quantitative transient stability assessment. Furthermore, the adverse impact of insufficient damping of GFM inverters on system transient stability is thoroughly investigated. The effectiveness of the proposed method is validated on the modified IEEE 14-bus system with mixed GFL and GFM inverters, and the numerical simulation results further demonstrate the small conservativeness of the proposed method.
Read the full abstract ▾
Social and Information Networks
Tracking and Predicting Evolution of Social Communities
Mark Goldberg, Malik Magdon-Ismail, Srinivas Nambirajan et al. (+1 more)
We develop an algorithmic framework for studying the
evolution of communities in social networks. We begin with the theoretical foundation, from which we conclude that an
evolution is at most as strong as its weakest link. This allows us to formulate an efficient
algorithm to identify all evolutionary sequences in a dynamic social network. We use this
algorithm to empirically study community
evolution in several large social networks, to identify those features of the early stages of a community that indicate whether a community is going to be shortlived or not. Our results show that it is possible to correlate the lifespan of a community to structural parameters of its early
evolution; these conclusions are robust across all the social networks we have investigated.
Read the full abstract ▾
All authors Mark Goldberg, Malik Magdon-Ismail, Srinivas Nambirajan, James Thompson
Computer Vision
Zekun Zhang, Yixiang Cai, Yuxi Liu et al. (+9 more)
Diffusion Transformers (DiTs) achieve strong
video generation quality, but their dense spatiotemporal self-
attention scales quadratically with sequence length and quickly becomes the dominant inference bottleneck. Sparse low-rank hybrids alleviate this cost by combining a local sparse branch with a global compressed branch. In
video DiTs equipped with 3D Rotary Position
Embeddings (RoPE), the global branch faces a structural compatibility issue: when RoPE is applied before a nonlinear feature map, the rotation and nonlinearity generally do not commute, making it difficult to keep a query-independent linear summary while preserving relative rotary geometry. Existing work often sidesteps this issue by replacing genuine cross-
token global aggregation with coordinate-conditioned surrogates or learnable absolute positional modules. These compromises can be effective, but they approximate relative decay from absolute coordinates and introduce extra positional parameters. We propose RoLA, a rotary-positioned low-rank linear-
attention branch that keeps genuine cross-
token aggregation while remaining compatible with a reusable linear summary. The design applies RoPE outside the nonlinear low-rank feature map and reuses a truncated subset of the pre-trained rotary schedule matched to the low-rank bottleneck. This yields a linear-time low-rank global branch with relative positional behavior by design and no additional positional parameters; the full sparse–low-rank module still includes the fixed-sparsity sparse branch. Experiments on open-source
video DiTs show that the resulting method remains competitive in generation quality at 90% sparsity while achieving 2.63$\times$ end-to-end inference speedup on Wan2.1-14B (720p, 81 frames, measured on an NVIDIA H100 GPU).
Read the full abstract ▾
All authors Zekun Zhang, Yixiang Cai, Yuxi Liu, Tengxu Sun, Tianle Liu, Zhoutong Wu, Haoyu Li, Baole Ai, Ang Wang, Jiamang Wang, Lin Qu, Kun Yuan
Cryptography and Security
A Trustworthy Watermarking Framework for LLM-Generated Food Safety Content
Zhongli Fang, Yiran Chen, Lingyun Zhang et al. (+4 more)
Large
language models are transforming many industries with their text generation abilities. However, their outputs can be easily tampered with, creating serious risks in critical areas such as food safety reporting. To protect the integrity and traceability of AI-generated content, this paper introduces ToSS (
Token Oriented Repartitioning and Strategic Selection), a reliable authentication method using adaptive dual watermarking. The key innovation of ToSS is its dual watermark encoding approach that divides vocabulary
tokens into black and white sublists, enabling precise bit-level
embedding of traceability information. Additionally, an
entropy adaptive mechanism dynamically selects text regions with high prediction uncertainty for watermark insertion, maintaining text fluency and factual accuracy while ensuring reliable traceability. Experiments on multiple
datasets, including food domain texts, demonstrate that ToSS achieves leading performance in both watermark capacity and decoding accuracy.
Read the full abstract ▾
All authors Zhongli Fang, Yiran Chen, Lingyun Zhang, Yu Liu, Ping Chen, Xiaoyan Sun, Jun Dai
Computer Vision
Counterfactual Tests for Measuring Chain-of-Thought Faithfulness in Visual Language Models
Bayar Menzat, Maximilian Süss, Ruizhi Wang et al. (+3 more)
Chain-of-thought (CoT) may often look plausible, yet it may not faithfully reflect the model's decision-making process. While methods for measuring the faithfulness of CoTs for textual inputs have been increasingly introduced, using these methods for visual inputs is not straightforward. In this work, we adapt the family of counterfactual methods for measuring CoT faithfulness, namely the Counterfactual Test (CT) and Correlational Counterfactual Test (CCT), to visual inputs, and call them vCT and vCCT, respectively. Using vCT and vCCT, we
benchmark eight recent open-source
Vision Language Models (VLMs) on two
datasets. Our analysis shows that CoTs do not reliably track visual evidence that influences model predictions: they may omit the removed object even when its removal causes a large prediction shift, yet mention it when the shift is small. We further find that Predict-then-Explain explanations align more strongly with perturbation-induced probability shifts than pre-answer CoTs, while binary vCT scores are often nearly saturated. We also include a reconstruction control, in which images pass through the same editing pipeline without object removal, and find that the main object-removal intervention induces larger shifts than reconstruction alone. We construct and release Counter-SNLI-VE and Counter-A-OKVQA, two
datasets of image pairs that differ by a single object.
Read the full abstract ▾
All authors Bayar Menzat, Maximilian Süss, Ruizhi Wang, Benno Steinegger, Thomas Lukasiewicz, Oana-Maria Camburu
Computation and Language
DianShi-RxnDB: A Large-Scale, Fine-Grained Organic Reaction Data Platform Built via a Fully Automated Pipeline for Researchers and AI Agents
Yubin Wang, Xingjian Wei, Jiang Wu et al. (+33 more)
High-quality structured organic reaction data are essential for developing artificial intelligence for chemistry (AI4Chem), yet much of this knowledge remains dispersed across patent text, images, and reaction schemes. We present DianShi-RxnDB, a large-scale, fine-grained organic reaction data platform built via a fully automated extraction and normalization pipeline integrating patent text, images, and reaction schemes. Its corpus covers organic synthesis patents from the USPTO and EPO published between 1976 and 2025, yielding approximately 24 million reaction instances, of which approximately 14.8 million (61.7%) pass automated qualification checks. Each instance represents a specific single-step experiment recording participants, roles, quantities, temperatures, reaction times, yields, experimental procedures, and provenance links to source patents. In a manual evaluation of 1,300 sampled qualified instances, the micro-averaged field-level accuracy was 92.95%. A matched comparison with Pistachio further indicated advantages in deduplicated record counts, representation granularity, and field-level exact agreement. The platform provides a Web research workbench for searching, filtering, comparing, and source-verifying records, and a Model Context Protocol (MCP) service offering AI
agents composable structured
retrieval tools. DianShi-RxnDB is available at https://dianshi.opendatalab.org.cn/ .
Read the full abstract ▾
All authors Yubin Wang, Xingjian Wei, Jiang Wu, Yinfan Wang, Boyu Zhu, Lin Zhang, Jianing Yu, Huazheng Zeng, Ruiyi Ding, Junyuan Gao, Jiaxing Sun, Lingli Ge, Haote Yang, Jingchao Wang, Aijia Guo, Qian Jiang, Yurui Zhao, Wenjian Zhang, Chen Zhu, Lijun Wu, Xiaolei Yang, Haodong Chen, Junjie Yuan, Zichao Ye, Shaowei Hou, Jing Ye, Jia Yu, Shan Wang, Lijun Wu, Jiantao Qiu, Chao Xu, Yuqiang Li, Guangyu Wang, Bowen Zhou, Dahua Lin, Conghui He
Computation and Language
PARSER: Read in Parallel, Reason in Depth for Long-Context LLM Agents
Kun Li, Zexuan Qiu, Tianhua Zhang et al. (+2 more)
Sequential
memory agents process long documents by reading chunks one after another while maintaining a compact
memory state, coupling document traversal to
reasoning depth. This coupling introduces sensitivity to evidence placement and ties inference latency linearly to document length. We introduce PARSER, which decouples reading from
reasoning. A bank of lightweight subagents each bound to a single chunk read the entire document in parallel, while a lead
agent reasons in depth through iterative scatter–gather rounds: at each round it broadcasts a query to all subagents, aggregates the returned evidence, and formulates a deeper follow-up query conditioned on what has been found so far. This decoupled design concentrates all learnable behavior in the lead
agent, which is optimized with reinforcement learning, while the subagents remain frozen off-the-shelf models. On multi-hop QA with contexts ranging from 7K to 896K
tokens, PARSER with a 4B backbone outperforms the strongest sequential
memory baseline by 5.7 points on average and by 12.0 points at 896K
tokens. Scaling to a 9B backbone, PARSER surpasses DeepSeek-V4-Pro by 6.3 points. Controlled experiments confirm that PARSER is robust to perturbations in evidence position, order, and distance, conditions that cause large accuracy swings in sequential methods, while reducing inference latency by up to 11x.
Read the full abstract ▾
All authors Kun Li, Zexuan Qiu, Tianhua Zhang, Irwin King, Helen Meng
Computer Vision
Attention-Enhanced Deep Features with Heterogeneous Ensemble Learning for Glaucoma Detection
Abdullah Al Shafi, Nishat Sadaf Lira, Abrar Hasan et al. (+2 more)
Glaucoma is a progressive optic neuropathy characterized by irreversible damage to the optic nerve, making timely diagnosis critical to prevent permanent
vision loss. Although deep learning has demonstrated promising performance in automated glaucoma detection, existing approaches often overlook feature refinement, suffer from class imbalance, and rely on individual classifiers that limit prediction robustness. To address these challenges, this paper proposes a hybrid glaucoma detection framework that integrates
attention-enhanced deep feature extraction with heterogeneous ensemble learning. Specifically, deep representations are extracted using InceptionV3 and subsequently refined by incorporating the Convolutional Block
Attention Module (CBAM) to enhance discriminative retinal features. To improve classification robustness, the extracted features are classified using multiple machine learning models together with Single-Level Ensemble (SLE) and Double-Level Ensemble (DLE) strategies, while SMOTE combined with Tomek Links (SMOTE+TL) is employed to alleviate class imbalance. Furthermore, a systematic comparison of handcrafted, deep, and
attention-enhanced deep feature representations is conducted. Experimental evaluation on two public retinal fundus
datasets demonstrates that deep feature-based methods consistently outperform handcrafted feature-based methods, while the proposed
attention-enhanced framework achieves the best overall performance. Furthermore, Grad-CAM visualizations confirm that the proposed model focuses on clinically relevant retinal regions, providing interpretable evidence on the model's prediction process.
Read the full abstract ▾
All authors Abdullah Al Shafi, Nishat Sadaf Lira, Abrar Hasan, Kazi Saeed Alam, Swapnil Kundu Argha
Systems and Control
User-specified Kelvin-hour budgets within model predictive control for energy-efficient buildings: Simulation and field demonstration
Jicheng Shi and Colin N. Jones
For energy-efficient building control, model predictive control (MPC) has been widely proposed to reduce energy use while maintaining prescribed indoor temperature bounds. In practice, MPC implementations may still produce temperature-bound violations because of model mismatch, weather-forecast errors, and softened constraints introduced to preserve feasibility. Existing soft- constrained MPC usually handles this trade-off through slack-weight tuning, so users cannot prescribe a violation-severity budget before operation. This paper develops a Kelvin-hour (Kh)-budget MPC that allows users to specify a running-average budget for temperature-bound violation severity. During operation, the controller uses the realized Kh budget surplus or overspend to update the temperature bounds, providing step-by-step budget feedback. We evaluate the method in two high-fidelity BOPTEST simulation cases and in an occupied residential deployment. In the first one-zone case, prescribed budgets produce clear running-average Kh violation responses across predictors and disturbance settings. At a prescribed budget of 0.005 Kh/step, the controller reduces energy use by 31.9% and Kh violation by 36.2% relative to the built-in default controller. In the second coupled two-zone case, separate zone budgets produce zone-level responses, with 18.9%-20.5% energy reduction and 9.4%-64.3% zone-level Kh violation reduction. In the occupied residential deployment, the running-average Kh violations remain close to or below their prescribed budgets under real sensing, actuation, weather, and occupancy conditions.
Read the full abstract ▾
Systems and Control · Cross-listed from stat.ME
Bayes Estimators with Performance Comparable to Empirical Bayes Estimators and Improved Local Robustness
Yue Ju, Ying Wang, Jiabao He et al. (+2 more)
Bayes estimation has been extensively studied and widely used in statistics, decision theory, signal processing, machine learning, and system identification. Among its variants, empirical Bayes (EB) estimation has attracted considerable
attention due to its favorable estimation performance and computational tractability. However, the direct plug-in dependence of an EB
estimator on hyperparameters can make it locally sensitive to hyper-parameter perturbations. This paper considers the linear regression model and focuses on the EB
estimator by employing the marginal maximum likelihood hyper-parameter
estimator. For conciseness, this
estimator is simply referred to as the EB
estimator. Given a family of EB weighting functions, a generalized Bayes
estimator is constructed with the same excess mean squared error (XMSE) as the corresponding EB
estimator. Here, the XMSE is a second-order asymptotic measure of the mean squared error difference between the
estimator of interest and the maximum likelihood
estimator. Furthermore, the EB
estimator is shown to be at most firstorder sensitive to hyper-parameter perturbations, whereas the constructed Bayes
estimator is at most second-order sensitive, making it locally more robust. The computational complexities of these two
estimators are also analyzed. In some cases, the constructed Bayes
estimator can be computationally comparable to, or more efficient than, the EB
estimator. These theoretical results are further supported by numerical simulations.
Read the full abstract ▾
All authors Yue Ju, Ying Wang, Jiabao He, Bo Wahlberg, Håkan Hjalmarsson
Distributed Computing
ForgeStencil: Automating Per-Case Stencil Specialization from Kernels to 100+ Real Applications
Yaojian Chen, Yuxuan Li, Wubing Wan et al. (+3 more)
Industrial and scientific computing rests on a few core kernels, and the stencil is among the most widely used: weather and
climate models, seismic imaging, fluid dynamics, and image processing all run on it. No single stencil implementation is fastest: the optimal kernel changes qualitatively with stencil shape, grid shape, precision, and host application. For two decades the field has answered with general methods (DSLs, code generators, autotuners), because specialized solutions were too expensive to build per case, so all reuse one human-authored recipe. That reuse costs performance; we call the cost the generality tax. This premise no longer holds: code-synthesis
agents now build a correct, specialized solution per case at acceptable cost. ForgeStencil automates this. A Kernel
Agent synthesizes CUDA and forges a per-configuration map of specialized operators, removing the tax case by case. On an A100 the map beats the strongest public baseline in 37 of 37 cases:
geometric mean 2.35x against same-precision f32 baselines and 1.95x for fp16, each reported under its own precision. The same change reaches end-to-end application performance. A generic operator library is tuned once for its own general case and reused across applications, so its shapes, layouts, and launch boundaries are optimal for none of them: using it is the application-level form of the tax. An App
Agent instead forges a specialized solution per application, locating hotspots, rewriting application structure, and validating and integrating each change. Across 100 real industrial and scientific codes the end-to-end median speedup is 1.41x against each application's own GPU baseline. To our knowledge this is the first demonstration that per-case synthesis carries from a kernel library to complete applications at this breadth, and evidence that reuse is no longer the default in a domain built on it for two decades.
Read the full abstract ▾
All authors Yaojian Chen, Yuxuan Li, Wubing Wan, Lin Gan, Guangwen Yang, Zhiyuan Liu
Distributed Computing
Gutenberg: Taming Latency-Critical Cloud Services with Near-Data-Processing
Qiushi Lin, Phillip B. Gibbons, Jovan Stojkovic et al. (+1 more)
Latency-critical cloud services place growing pressure on
memory while requiring isolation,
fairness, and predictable QoS. Near-data processing (NDP) reduces data movement by executing requests close to
memory, and prior systems further improve locality through caching and replication. However, writes make replica maintenance expensive, while uneven compute and
memory traffic can overload a few NDP units and increase tail latency. Existing throughput-oriented schedulers do not fully address these challenges for co-located cloud services. We present Gutenberg, a CPU+NDP for mutable, latency-critical cloud services. Gutenberg stages subpage updates in a CPU-resident delta buffer, allowing hot writable pages to remain replicated without eager full-page synchronization. It also adopts CPU helper cores to assist request execution when NDP execution or replica maintenance becomes costly. An online controller jointly decides page placement, replication, CPU/NDP execution, and routing using access patterns, queue pressure, and feedback from prior decisions. The system further enforces isolation and fair resource allocation across services. We also model-check CPU–NDP coordination protocol for correctness. We evaluate on TailBench using ZSim with Ramulator-calibrated
memory timing. Across evaluated services, Gutenberg outperforms prior systems, reducing average and p99 latency by up to 80.4% and 85.8%. It also improves isolation and
fairness while adapting to changing workload behaviors.
Read the full abstract ▾
All authors Qiushi Lin, Phillip B. Gibbons, Jovan Stojkovic, Yiwei Zhao
Computation and Language
A Grapheme-Aware Indic Tokenizer for Tamil: Large-Scale Training and Intrinsic Evaluation
Hari Krishnan K V and Sudarsun Santhiappan
Tokenization forms the foundation of modern Natural
Language Processing (NLP) systems by transforming raw text into discrete units that neural
language models can process. The effectiveness of this process directly influences vocabulary efficiency, sequence length, computational cost, and downstream model performance. Although
multilingual tokenizers such as Byte Pair Encoding (BPE), WordPiece, and SentencePiece have performed well across numerous
languages, they often segment morphologically rich Indic
languages inefficiently. Tamil, in particular, poses unique challenges because its grapheme-based writing system can represent a single visible character with multiple Unicode code points. In this work, we present a grapheme-aware Indic tokenizer for Tamil that preserves complete grapheme clusters through a reversible Unicode mapping strategy prior to WordPiece vocabulary learning. By operating on grapheme-level representations instead of individual Unicode code points, the tokenizer produces linguistically meaningful
token boundaries while remaining fully compatible with
transformer-based
language models. The tokenizer is trained on a large-scale Tamil corpus and evaluated using a comprehensive intrinsic evaluation framework that measures compression efficiency,
token fragmentation, information density, and vocabulary utilization. Experimental evaluation compares the proposed tokenizer against five widely used
multilingual tokenizers: GPT-2, mBERT, mT5, mBART, and NLLB. The proposed tokenizer achieves the strongest performance among the evaluated tokenizers on fragmentation- and sequence-efficiency-oriented intrinsic metrics, while matching the highest observed compression ratio. These results demonstrate the effectiveness of grapheme-aware preprocessing for Tamil tokenization.
Read the full abstract ▾
Machine Learning · Cross-listed from hep-ex
Likelihood-Based Unsupervised Anomaly Detection in CMS Dijet Events
Bhavishya Chebrolu, Hitesh Rasineni and Prajwal Aaryan Immadi
We present an unsupervised search for anomalous dijet events in
proton–
proton collision data using neural spline flow density estimation. A normalizing flow model is trained on a high-dimensional feature space comprising jet, dijet, and event-level observables to learn the dominant Standard Model background directly from data, without assuming a specific signal hypothesis. Events assigned low likelihood under the learned density are identified as potential anomalous events. Using this approach on a CMS Open Data dijet sample, we investigate extreme events in the tail of the anomaly-score distribution and perform an extensive validation and robustness study. This study includes feature-level statistical comparisons, mass decorrelation tests, permutation-based null tests, and evaluations of training stability. The selected anomalous events exhibit notable departures from the background-only expectation, primarily in jet-substructure observables, while remaining stable under several known sources of bias in unsupervised learning. The identified anomalies are distributed across the kinematic phase space and do not exhibit a narrow structure in the dijet invariant-mass spectrum. Instead, they show correlated deviations across multiple observables, consistent with a multivariate difference in jet substructure and event
topology rather than a localized resonance. While no claim of new physics is made, this study demonstrates that neural spline flow-based density estimation can be sensitive to rare, structured deviations in
collider data and may provide a model-independent exploratory tool for searches for physics beyond the Standard Model at the LHC.
Read the full abstract ▾
Machine Learning · Cross-listed from stat.ML
RoPE attention is an exact forward-pass gradient step with softmax intact
Julie Huang, Maggie Chlon and Leon Chlon
We derive an exact gradient-step representation of the RoPE-softmax forward pass. For every deterministic RoPE-softmax
attention head with arbitrary affine projection weights, we construct a query-dependent effective
matrix $\Delta M_i$ satisfying $y_i = \mu_i + u_i^\top \Delta M_i$, where $\mu_i$ is the uniform mean of the attended values and $u_i$ is the augmented query input. The construction applies the classical exponential divided difference $\rho = \phi_1$ to retain the softmax exactly. Its positive coefficients give a unit gradient-step representation on a query-conditioned quadratic objective. The same function connects the RoPE generator to exact positional finite differences. We derive a tokenwise formula for the error of reusing one query's
matrix and prove that a nonconstant finite-cache head cannot admit a globally exact affine query readout. Reconstruction checks and frozen-reuse calibration on one pretrained Qwen2.5-0.5B layer verify the representation and quantify the correction required when one query's
matrix is reused.
Read the full abstract ▾
Numerical Analysis
A discontinuous Galerkin approximation of the Dean–Kawasaki equation
Kamran Arora, Federico Cornalba and Tony Shardlow
We introduce and analyse an arbitrary order spatial discontinuous Galerkin (dG) method for the Dean–Kawasaki equation, a highly singular SPDE modelling density fluctuations of $N$ diffusing particles in the regime of large particle number $N \gg 1$. Our starting point is a general procedure for discretising multiplicative, divergence-form noise on finite element spaces whilst preserving its cross-variation structure at the discrete level; the construction is explicit, elementwise, applies to continuous and discontinuous spaces alike, and extends to general mobilities. Using it, we prove weak error estimates of order $O(h^p)$ between fluctuations of the semi-discrete scheme and those of the underlying particle system, together with a correction that is exponentially small in the scaling regime $Nh^d \gg 1$ and arises because the scheme does not preserve positivity. The resulting method is locally and globally conservative and applies on unstructured simplicial meshes. Quantitative numerical experiments support the analysis, and further experiments illustrate the method beyond the scope of the theory: indicator function observables, external and interaction potentials, convection-dominated regimes and reflecting boundary conditions.
Read the full abstract ▾
Distributed Computing · Cross-listed from physics.comp-ph
A HIP-Compatible Accelerator Backend for Fourier-Bessel Particle-in-Cell Simulations on CPU/DCU Heterogeneous Clusters
Jingliang Fan, Ruiqing He, Yang Wan et al. (+3 more)
FBPIC (
Fourier-Bessel particle-in-
cell) is a high-performance simulation code for relativistic
plasma and accelerator physics. Its original accelerator backend relies on Numba CUDA, which limits its direct deployment on accelerators using the HIP (Heterogeneous-Compute Interface for Portability) programming environment, such as DCU (Deep Computing Unit) accelerators. In this work, we develop an accelerator backend compatible with HIP that enables FBPIC to run efficiently on DCU platforms while preserving its Python user interface and high level simulation workflow. For the evaluated LWFA (
laser-wakefield acceleration) workloads, the proposed backend achieves 1.32-1.54x speedups over the original FBPIC implementation on an NVIDIA V100 GPU and enables efficient execution on the DCU platform. We also summarize the key lessons learned from porting FBPIC to the DCU platform. Multi-DCU experiments achieve a 1.88x strong-scaling speedup on four accelerators and a 2.72x increase in aggregate throughput at approximately 68% weak-scaling efficiency, with communication analysis identifying inter-node communication and synchronization as the main scalability limitations. Beyond FBPIC, the proposed approach provides a practical reference for porting and optimizing other scientific computing applications developed with Python on heterogeneous accelerator platforms.
Read the full abstract ▾
All authors Jingliang Fan, Ruiqing He, Yang Wan, Jiandong Shang, Hengliang Guo, Qiang Chen
Cryptography and Security
Detokenization Leaks: Reconstructing Local LLM Outputs From Cache Traces
Roy Weiss, Benyamin Konstantinov, Eitam Sheetrit et al. (+2 more)
We present a new attack that reconstructs the text generated by locally hosted LLMs by observing CPU cache activity during detokenization. Unlike prior attacks that rely on deployment-specific assumptions, such as shared data
memory, CPU offloading, or Mixture-of-Experts architectures, our approach targets the detokenizer, a component used in default LLM inference pipelines. To obtain clean signals, we use Flush+Reload on shared tokenizer code to detect when decoding occurs, which lets us perform
Prime+Probe at the right moment and isolate
token-dependent cache activity. We then apply a clustering-and-
language-model pipeline to recover text from noisy cache observations. We evaluate the attack across multiple
datasets, hardware platforms, inference frameworks, and model families, and show that it can recover semantically accurate outputs from real-world local LLM deployments, including agentic systems. This vulnerability is particularly significant because the most widely used tokenizer implementations are susceptible to the attack and are embedded in many popular local LLM products and
agent frameworks, including systems such as OpenClaw (which we demonstrate), substantially broadening the practical attack surface.
Read the full abstract ▾
All authors Roy Weiss, Benyamin Konstantinov, Eitam Sheetrit, Tomer Simon, Yisroel Mirsky
Machine Learning
Tracking the Moving Frontier: Long-Short Term Advantage Estimator
Xinhao Yao, Lu Yu, Changhao Wang et al. (+5 more)
Group-based RLVR methods estimate advantages by repeatedly sampling multiple trajectories for each
prompt, making long-horizon
agent training expensive and discarding useful experience accumulated across iterations. We ask whether historical experience can replace these repeated within-iteration comparisons without directly optimizing on stale trajectories. We introduce Long-Short Term Advantage
Estimator (LSTAE), a single-stream RL
algorithm that uses history for advantage estimation while updating the policy only with the current rollout. LSTAE maintains a persistent tracker for each task anchor. At the trajectory level (long term), a drift-aware historical baseline tracks the anchor's moving success frontier and measures the relative contribution of each new trajectory. At the step level (short term), a recent state-experience buffer exploits recurrent states to estimate localized action advantages. This two-timescale design converts accumulated experience into multi-granular credit signals, requiring only one rollout per anchor. Across agentic and mathematical
reasoning benchmarks, LSTAE matches or improves upon strong group-based baselines while substantially reducing rollout cost.
Read the full abstract ▾
All authors Xinhao Yao, Lu Yu, Changhao Wang, Fengwei Teng, Yuyao Zhang, Qing Cui, Jun Zhou, Yong Liu
Machine Learning
Data Efficient Sample Selection for In-Context Learning
V Venktesh, Cem Levi and Avishek Anand
The In-context learning (ICL) paradigm aids large
language models (LLMs) to adapt to new tasks without need for fine-tuning. However, selecting an optimal combination of demonstration examples from a large pool of example subsets is a challenging problem. Existing approaches for selection do not model the complex relationship between ICL samples and downstream LLM performance. They typically perform static task-level selection, choosing subsets once offline, which can fail to generalize to unseen queries. We introduce DearICL (Data Efficient
Algorithm for Ranking) ICL samples, a new framework that models demonstration example selection as a subset ranking problem. DearICL employs a non-linear surrogate employing a differentiable sorting objective within a gap-index bandit
algorithm. The gap-index based approach enables fine-grained separation of good arms and borderline arms, which is used as an auxiliary objective to train the non-linear surrogate through sufficient sampling of borderline arms, supporting instance-level subset ranking. On exemplar selection
benchmarks with open-source LLMs, DearICL achieves 8.08-15.9% accuracy gains over strong linear bandit baselines, with low sample complexity. Code and data: https://github.com/VenkteshV/DearICL.
Read the full abstract ▾
Data Structures and Algorithms
Tight Bounds on the Cost of Adaptivity for the Meyerson Sketch
Edith Cohen, Elena Gribelyuk, Pasin Manurangsi et al. (+1 more)
In online facility location, points arrive one at a time, and the
algorithm must either open a facility at the arriving point or route the point to an existing facility. The Meyerson sketch opens a facility at each arriving point with probability proportional to the point's distance to the closest open facility, and requires no state beyond the set of open centers. Due to its simplicity, space efficiency, and strong guarantees against the offline optimum, the Meyerson sketch has become a workhorse of streaming and online clustering. In many such applications, however, the set of open facilities are visible to the process that generates the stream, which can adaptively select future points based on the
algorithm's past random choices, voiding its classical guarantees. In this work, we quantify the effect of such adaptivity. We compare an adaptively generated run of the sketch against an oblivious replay, an independent execution, with fresh coins, on the very same generated sequence, and study the adaptivity ratio of expected adaptive cost to expected replay cost. We determine the worst-case ratio in both directions: adaptivity can neither inflate nor deflate the expected cost, or the number of open facilities, by more than an $O(\log\Delta/\log\log\Delta)$ factor, where $\Delta$ is the aspect ratio of the input points (the ratio of the largest to the smallest pairwise distance). This is asymptotically tight as there are deterministic generators on the real line that inflate or deflate the cost by an $\Omega(\log\Delta/\log\log\Delta)$ factor. We show that these robustness guarantees carry over to Meyerson-based sketches for approximate $k$ clustering with sketch size $O(k\,\mathrm{polylog}(n))$.
Read the full abstract ▾
All authors Edith Cohen, Elena Gribelyuk, Pasin Manurangsi, Uri Stemmer
Machine Learning
Towards Unified Multimodal Graph Foundation Model: A Bridge-Router-Adapter Based Approach
Sirui Zhang, Yubing Zhou, Xunkai Li et al. (+6 more)
Multimodal graphs couple node attributes in different modalities, such as text and images, with relational structure, enabling topological structure and cross-modality attributes to be modeled jointly.
Multimodal graph foundation models seek unified representations from such data that transfer across different graph domains and downstream tasks. However, existing methods exhibit two fundamental limitations. (1) Cross-Scope Context
Entanglement. They merge scope-specific graph contexts into a unified representation, obscuring their distinctions during
multimodal construction. (2) Scope-Ignorant Modality Routing. They route modalities within a fixed graph scope, overlooking how modality relevance varies across neighborhood ranges. To address these challenges, we propose
BRAIN, a unified model that focuses on graph context that combines neighborhood scope with modality composition.
BRAIN comprises a scope-conditioned Bridge that combines structural information spanning local-to-global neighborhood scopes with different modality compositions; a hierarchical Router that estimates the relevance between the scope and the task, and selects compositions separately within each scope, allowing modality utility to vary with graph range; and a lightweight residual Adapter that further specializes the routed
embedding for downstream prediction.
BRAIN is trained through multi-graph pretraining followed by task-specific adaptation. Experiments across nine
datasets and four task families demonstrate its broad effectiveness, improving node-classification and link-prediction performance by up to 4.73% relative to the strongest baseline, while achieving an average relative improvement of 14.72% across four graph-to-text and two graph-to-image metrics.
Read the full abstract ▾
All authors Sirui Zhang, Yubing Zhou, Xunkai Li, Zekai Chen, Shumeng Li, Wang Luo, Yinlin Zhu, Yujin Gao, Rong-Hua Li
Machine Learning
Behavioral Cloning Outperforms Entropy-Regularized RL: Critic-Driven Failure of Actor-Critic Methods on Adaptive Tumor Treatment
Aleksandar Dimitrov and Giacomo Spigler
Adaptive dosing requires policies that reduce tumor burden without excessive toxicity. Learned dosing policies are typically judged against historical or heuristic comparators, which cannot show whether a policy has found the best behavior available. We instead study a three-population tumor-control ODE in which optimal-control analysis fixes the form of a good schedule – bang-bang dosing punctuated by a singular arc – and construct a numerical controller of that form as a proxy for near-optimal behavior. Judged against this reference under a sustained-cure criterion – 200 consecutive days below 5% carrying capacity – Soft Actor-Critic (SAC) trained from scratch never reaches cure. Behavioral cloning (BC) of the reference reproduces it (100% sustained cure, 30/30 seeds), but SAC fine-tuning of the cloned policy destroys it across five
entropy coefficients, and TD3 and BC-regularized SAC fail identically; the pattern persists under multiplicative pharmacokinetic action noise. Along curative trajectories the post-collapse critic ranks the collapsed-policy action above the reference action in 96% of states, concentrated in the maintenance phase, and the policy settles into a non-curative adaptive-therapy
equilibrium. The reference is what makes this legible: against a heuristic comparator the fine-tuned policy would read as a competent controller rather than a failure.
Read the full abstract ▾
Computer Vision
TransNormal-2: Geometry-Grounded Rectified Flow with Edge-Aware Decoding for Precise Normal Estimation
Mingwei Li, Yi Yang and Hehe Fan
Diffusion-based models enable
monocular geometry estimation, yet their pixel-space precision is limited by a shared, under-studied error source: VAE reconstruction degradation. The 8x spatial compression in the VAE encoder-decoder degrades surface normals at object boundaries; even encoding and decoding ground-truth normals introduces 1.3–8.5{\deg} of mean angular error (MAE), with edge MAE reaching 2.8x the global MAE. We present TransNormal-2, a FLUX.2-based rectified-flow framework with single-step deterministic inference that addresses this degradation on both sides of the VAE decoder: in how latent predictions are supervised during training, and in how decoded normals are corrected at inference. First, geometry-aware pixel-space losses, including inverse rendering self-consistency, von~Mises-Fisher angular loss, and wavelet edge-aware regularization, complement latent MSE by enforcing spherical normal geometry and diffuse image-formation cues after VAE decoding. Second, a lightweight
Geometric Refinement Module (GRM) applies an RGB-guided residual correction to reduce boundary-localized decoding errors without freely rewriting the coarse prediction. On general-scene
benchmarks, TransNormal-2 matches or exceeds MoGe-2 on all eight reported metrics while using only 1.4% as many task-specific normal annotations. The gains are clearest for transparent objects, reducing MAE by 4.2{\deg} on ClearGrasp and 3.1{\deg} on ClearPose over the strongest prior baselines. Code will be released at https://longxiang-ai.github.io/TransNormal-2.
Read the full abstract ▾
Information Retrieval
EviMap: Evidence-Grounded Hierarchical Topic Maps for Exploring Unlabeled Corpora
Zhiyin Tan and Changxu Duan
Research teams and organizations often explore unfamiliar free-text collections, from survey comments and reviews to reports and domain documents, before labels, queries or coding schemes exist. At this stage, the first thematic map shapes what users notice, prioritize and carry into downstream analysis, so it should be trusted only insofar as it can be verified. Existing options force a trade-off between scale and verifiability. Qualitative coding preserves evidence but is slow. Search presupposes a query. Clustering and topic models scale but produce labels users must interpret. One-shot large
language model (LLM) summaries are fluent yet difficult to reproduce or audit. We present EviMap, an interactive system providing researchers and practitioners with an auditable thematic overview of such corpora. Guided by model-generated context describing the corpus and hypothesized stakeholder concerns, EviMap extracts within-document evidence phrases and organizes them, rather than whole documents, into a three-level map of aspects, groups and fine-grained topics.
Embedding-based clustering narrows the search space for finer semantic judgments by the LLM. Each node traces back to supporting phrase spans, so documents link to topics through evidence they contain and users can audit labels against the original text. Users can start from a top-level corpus map, drill into topics, inspect highlighted evidence in original documents, and combine two topics to find documents discussing both. We demonstrate this workflow across six heterogeneous corpora spanning 2,108 to 101,699 documents, with a comparison against flat and hierarchical LLM baselines. By grounding every label in verbatim source spans, EviMap makes a topic map not just readable, but verifiable. Code, demo
video, and interactive dashboard are available at https://github.com/zhiyintan/EviMap.
Read the full abstract ▾
Computation and Language
ECOKV: Geometry-Aware KV Cache Eviction via Complementary Diversity Metrics
Chin Ting Hsu, Yu-Syuan Xu, Ling Zou et al. (+2 more)
Although
multimodal Large
Language Models (MLLMs) excel in diverse tasks, their scalability remains limited by the
memory and computational overhead of KV cache storage. Recent KV cache eviction approaches incorporate a cosine similarity-based diversity metric with importance metrics to selectively retain critical key-value pairs. However, cosine similarity involves normalization that discards magnitude information, and it often yields uniformly high similarity values across layers due to the anisotropy property of hidden representations. In our study ECOKV, we rigorously deconstruct the capabilities of existing diversity metrics. Moving beyond simple measurement, we propose a geometry-aware composite metric that jointly leverages Euclidean distance and cosine similarity to capture
token diversity from complementary perspectives. Furthermore, we use these two metrics to estimate the redundancy level of each
attention head, allowing adaptive weighting between diversity and importance scores during
token selection. Finally, we demonstrate that the observation window commonly employed to preserve recent
tokens can be substantially reduced, thereby allocating more cache capacity to informative
tokens and yielding consistent improvements. Extensive experiments demonstrate that ECOKV achieves state-of-the-art performance under various compression ratios and can be seamlessly integrated with existing KV cache eviction methods. We further analyze the relationship between importance and diversity, and examine redundancy patterns across layers and
attention heads.
Read the full abstract ▾
All authors Chin Ting Hsu, Yu-Syuan Xu, Ling Zou, Hsien-Kai Kuo, Wen-Huang Cheng
Human-Computer Interaction
The Effectiveness of Virtual Patient Simulation Versus Peer Simulation in Providing Sexual Counseling During Pregnancy: A Randomized Controlled Trial
Neslihan Yilmaz Sezer, Menekşe Nazl{\i} Aker, Pinar Kullu et al. (+3 more)
Although sexual health counseling is one of the important responsibilities of healthcare professionals, effective educational methods are needed to develop students' counseling skills in this field. The aim of this study was to compare the effectiveness of virtual patient simulation and peer simulation methods in developing sexual counseling skills during pregnancy among nursing faculty students. This randomized controlled trial included 51 participants assigned to one of three groups: virtual patient simulation, peer simulation in a virtual environment, or face-to-face peer simulation. In the study, all groups received face-to-face theoretical instruction on sexual counseling during pregnancy. Following the theoretical training, students participated in virtual patient simulation, peer simulation in a virtual environment, or face-to-face peer simulation practices according to the groups to which they were assigned by randomization. Outcome measures included the Sexual Attitudes and Beliefs Scale, the Student Satisfaction and Self-Confidence in Learning Scale, and the Sexual Counseling Skills Evaluation Form. It was determined that the participants' SABS scores decreased significantly after the intervention. According to the results of the mixed repeated-measures ANOVA, the effect of time was statistically significant. However, the group effect and the group x time interaction were not significant. There was no difference between the groups in terms of Satisfaction in learning, Self-confidence in learning, and Skill scores. This study showed that different simulation methods used in sexual counseling education during pregnancy were effective in reducing students' negative attitudes and beliefs and provided similar educational outcomes. Therefore, virtual patient and peer simulations may be recommended as feasible approaches for improving sexual counseling skills.
Read the full abstract ▾
All authors Neslihan Yilmaz Sezer, Menekşe Nazl{\i} Aker, Pinar Kullu, Rana Tuncer, Gael Lucero-palacios, Roghayeh Leila Barmaki