Machine Learning
Can Rényi Accountants Beat Lossless Accountants for DP-SGD?
Jinho Bok
Recent frameworks in differential
privacy (DP), such as
privacy loss distributions and $f$-DP, provide lossless representations of the
privacy leakage of mechanisms. These approaches are often considered to yield tighter $(\varepsilon, \delta)$-DP guarantees than those obtained from relaxed notions such as Rényi DP. But is that always the case? We study this question for differentially private
stochastic gradient descent (DP-SGD) with fixed-size batches under replace adjacency. Our key observation is that in this setting,
privacy accounting based on Rényi DP can use a pair of worst-case distributions (i.e., a dominating pair) that differs from the one used in the lossless frameworks. In particular, we characterize the worst-case distributions as two-dimensional Gaussian mixtures and derive an improved Rényi DP bound for DP-SGD. Numerical and asymptotic comparisons show that across various hyperparameter settings, a Rényi DP accountant based on the new bound can yield better
privacy guarantees than those obtained from the lossless frameworks.
Read the full abstract ▾
Statistics Applications
Spatial Partner Markets: Effective Competition, Scale, and Opportunity
Curren K. Mehta
Local partner
markets differ not only in population size but in relative partner availability, the scale of opportunity, marriage timing, and population turnover. This paper develops a county-level framework for comparing these conditions across local partner
markets. The empirical application focuses on unmarried women ages 25-34 and uses observed partner-age pairings to construct age-specific weights for the relevant unmarried male population. We model absolute opportunity separately, fitting a Michaelis-Menten diminishing-returns transformation to 192 scenarios that vary dating intensity, search horizon, fit, and first-date conversion. The resulting index also incorporates the share of women ages 35-39 who have never married and the inflow rate of adults ages 25-34. Among 320 counties with complete data, the District of Columbia ranks first and New York County second under the baseline specification, and rankings remain highly stable under alternative component weights. An endogenous mobility extension shows that favorable
markets attract additional male participants, producing substantial adjustment while leaving the top composite rankings relatively stable over the parameter values considered. The framework is intended as a structural comparison of local partner-
market conditions rather than a predictor of individual relationship outcomes.
Read the full abstract ▾
Artificial Intelligence
When LLM Agents See a Conflict but Still Proceed: Cross-Domain Dependencies in Network Slicing
Marcos Lima Romero and Ricardo Suyama
Future AI-native networks are being designed to apply autonomous
agents built on large
language models (LLMs) in network management through intents. When intents from different domains meet on the same resource, an
agent must recognize the cross-domain dependency before taking an irreversible action. We investigate this problem through a smart-grid use case, in which a power utility holds an active intent on a backup slice for protection traffic. This slice appears idle by design, which is precisely the profile targeted by a cost-reduction intent. We built a tool-based environment in which the dependency must be either queried by the
agent or read from an inventory annotation, together with a negative control, and evaluated ten models from four different providers in 600 runs. Most models acted against the dependency they had observed: seven of the ten models removed the protection slice at least once after receiving the cross-domain dependency, and in the high-harm tier the slice was removed in 76% of the exposed runs, while four models practically never did. Only one of ten models never removed the slice. We identify two failure modes, not querying the dependency and querying it but proceeding anyway, and show that
agents often reported no conflict after observing it. These results indicate that, in AI-native networks, visibility of cross-domain dependencies is not a sufficient safeguard, and that dependency verification should be enforced by the management architecture before irreversible actions rather than delegated to the LLM's judgment.
Read the full abstract ▾
Computation and Language
Gluing Local Contexts into Global Meaning: A Sheaf-Theoretic Decomposition of Transformer Representations
Bryce Grant and Peng Wang
Existing interpretability tools do not test whether a representation keeps the same content when only the phrasing changes. We treat each paraphrase pair as a local witness that two activations should agree and build a cellular sheaf over the paraphrase graph; its
Laplacian $L_\mathcal{F}$ answers the question on a frozen model with no training. The kernel $H^0$ is a content subspace: at 20 dimensions it beats the preserved subspace of the concept eraser LEACE on held-out CounterFact fact
retrieval across eight architectures by a mean $+17.1$\,pp, and constraining all paraphrases of a fact jointly adds a further $+6.5$\,pp. The spectral complement concentrates causal influence: ablating the most paraphrase-variable coordinates, a coordinate-wise proxy for it, changes model outputs $5.6$ to $26.5\times$ more than ablating coordinates of matched per-coordinate variability, across nine models from 124M to 13B parameters. Stacking layers into a grid creates $b_1 = N(L-1)$ algebraic 1-cycles, and the resulting model-derived harmonic mass varies by nearly three orders of magnitude across six architectures and ranks them as their steering fragility does (Spearman $\rho = 0.90$, exact one-sided permutation $p = 0.014$). The same paraphrase graph yields the content subspace, the influence directions, and the fragility measure. Project page: https://cwru-aism.github.io/gluing-lc-page/
Read the full abstract ▾
Classical Analysis and ODEs
Finiteness of Fuchsian equations with unitarizable monodromy on compact Riemann surfaces
Sheng Tan
We prove that, on any compact Riemann surface, there are only finitely many meromorphic projective connections with prescribed singular points and real exponent differences whose projective monodromy is unitarizable. We also establish properness and growth estimates for holonomy, including a relative version for smooth marked families, and derive finiteness results for spherical conical metrics and ramification divisors of stable extensions.
Read the full abstract ▾
Statistics Applications
Machine Learning Models Trained on a Longitudinal Survey Dataset Poorly Predict Hundreds of Life Outcomes
Emily M. Cantrell, Pranay Anchuri, Hanzhang Ren et al. (+1 more)
Understanding the predictability of life outcomes is important for public policy and science, but existing studies focus on a limited number of outcomes and therefore struggle to describe the broad patterns of predictability. We developed a high-throughput approach that uses a single
dataset to study the predictability of many future outcomes using past data. We deployed this approach with hundreds of outcomes and thousands of machine learning models per outcome in the Future of Families and Child Wellbeing Study. Three findings emerged. First, most outcomes were not very predictable, but some were much more predictable than others. Second, even for the most predictable outcomes, our complex prediction procedure typically exhibited little skill relative to a simple procedure. Third, we found modest patterns in how predictability varied across outcomes. We conclude by discussing scope conditions and how future research might address them.
Read the full abstract ▾
All authors Emily M. Cantrell, Pranay Anchuri, Hanzhang Ren, Matthew J. Salganik
Robotics
D-World: An Explicit World-Language-Action Model with Recursive Self-Improvement under Dual Control
Xuan Wen and Baichao Wang
Frontier large
language models (LLMs) now reason about space well enough to program
robots, yet the world models behind
robot policies are implicit, held in network weights or predicted
video, and cannot be inspected or corrected directly. We present D-World, an explicit world-
language-action model built on the Sense, Plan and Act decomposition. A perception
agent writes the scene as an executable digital twin using generated
vision and geometry code. A planning
agent searches for actions in simulation and, after each executed action chunk, revises the twin's object states and named physical parameters against what the
robot measured. A parameter revision is accepted only if replaying the latest chunk does not increase any measured residual beyond its tolerance, and every chunk is checked in the updated twin before it runs. Structural mismatches return to perception for reconstruction. We call this measurement-grounded cycle of correction and revision recursive self-improvement (RSI), in which the LLM stays frozen while the world it acts in improves. We model the choice of actions that both advance the task and expose model error as a dual control problem and show that the cost of incorporating evidence into the twin sets the threshold at which probing pays, so cheap local revision counteracts the turn-off of cautious controllers. Each episode runs a real-to-sim-to-real loop, and its recorded twins can be refitted and re-rendered offline into higher-fidelity environments for post-training. Because perception and planning communicate only through the twin and
language, either can be replaced by a learned model, and since inference dominates the loop, the loop speeds up with faster models. On a dual-arm
robot, towel and toaster tasks show local revisions, such as one reducing the bread's rendered-depth residual from 22.4 to 1.5 mm, that let execution continue without reconstruction.
Read the full abstract ▾
Computation and Language
Calibrated Typed Decisions for Entity, Relation, and Event Extraction
Yang Wang, Hanlin Xiao, Yanan Ma et al. (+2 more)
Information extraction models often perform well on the
datasets they are trained on, but transferring to new domains and annotation guidelines remains challenging. Supporting user-specified types alone does not resolve this problem. We present CalOn, a unified 0.6B-parameter encoder for entity, relation, and event extraction across
datasets and schemas. The model jointly encodes text and type definitions, then directly identifies entity spans, relations between spans, and event triggers and arguments. Its extraction decisions consider the requested types together, using their combined evidence to identify relevant candidates before assigning a type. This gives all three tasks a common probabilistic formulation while keeping every extracted mention anchored to the input text. We train on diverse sources, and supplement supervised learning with a reward that encourages correct extractions and penalises incorrect ones. Across 40
benchmarks, CalOn outperforms the evaluated baselines on most
benchmarks. It leads all reported relation extraction
benchmarks and achieves the highest average F1 for event triggers and arguments with gold triggers. The same model also performs strongly on aspect sentiment triplet extraction and pronoun resolution through its span-pair formulation. A correction for the probability bias introduced during training improves confidence estimates for entities, relations, and event arguments after calibration, with little change in F1.
Read the full abstract ▾
All authors Yang Wang, Hanlin Xiao, Yanan Ma, Chenghao Xiao, Chenghua Lin
Computational Geometry
The Rhythm of the Tiles: Narrow and Low Temperature Computation and Simulation in Synchronous Tile Self-Assembly
Florent Becker, Phillip Drake, Benjamin Durbin et al. (+3 more)
In this paper, we prove that the synchronicity provided by the syncTAM can be leveraged to perform two additional feats impossible in the asynchronous aTAM. The aTAM has long been known to be equivalent to finite automata in the one-dimensional case. For our first set of results, we prove that this still holds when a finite, fixed amount of space is available in the second dimension. Yet, surprisingly, in the synchronous setting, a small, fixed, width is enough to simulate a Turing Machine. This holds even in the non-cooperative setting. For our second set of main results we prove that there exists a single set of syncTAM tiles that can be used to simulate any syncTAM system simply by using those tiles to create a scaled version of the seed structure of the original system and then letting the tiles from that set self-assemble from the scaled seed structure. While it has been shown (FOCS 2012) that such a tile set also exists in the aTAM for simulating all aTAM systems, that "intrinsically universal" (IU) tile set requires cooperativity (and is impossible without cooperativity as shown in SODA 2014). We prove there exists an IU tile set in the syncTAM that works without cooperativity. Furthermore, there exists a tile set that is IU for the class of directed systems (i.e. systems which each have a single terminal assembly) while remaining directed during simulation. This was shown to be impossible in the aTAM (FOCS 2016).
Read the full abstract ▾
All authors Florent Becker, Phillip Drake, Benjamin Durbin, Nicolas Ollinger, Brent Orlina, Matthew J. Patitz
Optics
Chiral Vector Beams for Topological Enantio-Sensitive Imaging
A. Rodriguez, J. Edginton, A. Löhr et al. (+7 more)
Molecular
chiral discrimination is vital across chemistry and life sciences, yet optical approaches remain highly inefficient for small organic molecules. Here we turn vector beams, whose polarization carries a topological charge, into an efficient probe of molecular handedness. We show that the combination of a tightly focused radially polarized beam at the fundamental frequency {\omega} and an azimuthally polarized beam at 2{\omega} creates
chiral vector beams, a locally
chiral field with structured handedness. Using state-of-the-art computational modeling in the
chiral molecule carvone, we show that the nonlinear response (notably sum-frequency generation) of randomly oriented
chiral molecules to this
chiral vector beam captures the fields' local handedness and is highly enantio-sensitive. The two enantiomers emit light at frequency 3{\omega} with topological ring structures of different divergence, and thus the molecular handedness can be read out directly from the ring radius. The signal inherits the Poincaré index of the driving field, opening up new ways to robust, ultrafast, and and all-optical
chiral discrimination.
Read the full abstract ▾
All authors A. Rodriguez, J. Edginton, A. Löhr, J. Terentjevas, P. Vindel-Zandbergen, A. Ordóñez, J. Vogwell, O. Smirnova, D. Ayuso, L. Rego
Software Engineering
AIPOCH Open-Science: An Open-Source AI Research Workbench with Reproducible .science Packages
Wen Ouyang, Bocheng Huang, Qianyu Yao et al. (+76 more)
AI-assisted research leaves instructions, computations, artifact revisions and checks whose relationships matter when another researcher receives the work. We present AIPOCH Open-Science, an open-source research workbench with portable .science session packages, and specify an exchange contract for these recorded relationships. Version and assessment targets remain attributable while receiving-workspace identifiers can be remapped, eligible contents can be omitted explicitly, and imported history remains separate from a writable fork. A controlled session tested an incorrect first version, a reviewed correction, an unreviewed revision and a planned but unexecuted step. We compared complete and partial exports with source records extracted independently of the exporter, then inspected imports into separate projects within the same installation. Both conditions retained the tested version and review associations, kept the unreviewed revision distinct from earlier assessments and represented excluded content as unavailable. A documentation-rich ordinary directory also supplied evidence for all seven static inspection questions. A separate registry case connected a table to its recorded task, execution and review; offline execution of its retained script and fixed input reproduced two saved CSV tables byte for byte. The study contributes an implemented exchange contract and a controlled examination of its preservation rules, content-selection constraints and interpretation boundaries.
Read the full abstract ▾
All authors Wen Ouyang, Bocheng Huang, Qianyu Yao, Fei Sun, Minying Deng, Haoyue Cheng, Wenwen Zhou, Lei Mou, Yusi Wei, Jiang Li, Kun Zhu, Lei Wang, Yi Feng, Jiaheng Dai, Ke Qian, Dan Wang, Yahui Wang, Chenyi Wang, Mingsheng Lyu, Yaqiang Zhang, Ping Zheng, Chunhua Tang, Xinge Ma, Xu Zhang, Youjia Fan, Can Lu, Zhijun Dai, Man Hu, Qiqi Xie, Beinan Liu, Yuanning Guo, Yunze Liu, Jinhong Jiang, Xianjun Fu, Mingwei Zhang, Yuanji Xu, Shilin Xia, Bowen Yang, Mingjun Yu, Jingdun Xie, Xuwei Chen, Xia Zou, Yuzhou Xue, Jianguo Zhou, Chunhao Cao, Donghao Tian, Lei Huang, Jinxiao Li, Shengrong Long, Zefeng Song, Honglin Li, Jiang Luo, Wenying Jiang, Shanshan Li, Lei Dai, Ning Jiang, Ying Cai, Chaohui Li, Chunyan Peng, Hetong Zhao, Angqing Li, Chunyan Yang, Chongyang Sun, Yong Li, Xinyu Shi, Wei Dai, Ming Peng, Chen Sang, Yuanyuan Han, Desong Qiu, Haixin Miao, Jian Xu, Yi Chen, Xiaoliang Shi, Tian Shen, Teng Hu, Shuhao Wang, Yingyong Hou, Huimei Wang
Machine Learning
Masked Diffusion for Travel Diaries Synthesis
Pascal Jutras-Dubé, Raziq Zaman, Fabian Bastin et al. (+2 more)
Because household travel surveys combine household characteristics, variable-length person rosters, vehicle rosters, and trip chains, existing synthesis pipelines split them across separate models or flatten them into cropped fixed-length vectors. Representation, rather than modeling, becomes the bottleneck: splitting severs cross-segment dependence, and cropping truncates the household-structure distribution. We instead tokenize the diary and keep the object whole: each household-day becomes one
token sequence with count
tokens for vehicles, persons, and trips, and distinct symbols for padded, hidden, unobserved, and inapplicable positions. We introduce DiaryBridge, a masked
diffusion model that bridges the fully hidden sequence to the distribution of complete diaries by unmasking positions in any order. One trained model, conditioned on any observed subset, performs synthesis and partial-diary completion alike. Experiments on real data extracted from the 2017/2018 Regional Travel Survey of the Baltimore region show that DiaryBridge generates travel reflecting vehicle availability and employment status, recovers shared trips using other household members' observed diaries, and better reproduces aggregate demand across modes, purposes, and departure periods than the learned baselines.
Read the full abstract ▾
All authors Pascal Jutras-Dubé, Raziq Zaman, Fabian Bastin, Cinzia Cirillo, Ruqi Zhang
Biomolecules
Calibrating AlphaFold confidence scores for interface quality
Sagyndyk Sagadinov, Vsevolod Viliuga and Arne Elofsson
AlphaFold confidence scores are widely used to rank
protein complex predictions and to select subsets of models for downstream applications. Although these scores provide strong discriminatory power for distinguishing prediction quality, their numerical values were never intended to represent the probability that a prediction satisfies a given structural quality criterion. In this work, we calibrate commonly used confidence scores to estimate the probability that a predicted complex exceeds a chosen threshold of DockQ, an established measure of interface quality. On antibody-antigen and general multimer
datasets, post-hoc isotonic calibration consistently improves agreement between predicted probabilities and observed DockQ success rates while preserving the underlying score ordering. The calibrated probabilities also provide more accurate estimates of the fraction of selected models expected to pass a given structural quality threshold. Although the mapping is fitted within a given prediction distribution and therefore transfers imperfectly across prediction domains, we show that reliable estimates are recovered from a small number of labeled target-domain samples. Because the calibration is entirely post hoc and requires no underlying AlphaFold model fine-tuning, it provides a straightforward and customizable way to convert existing model quality rankings into calibrated probabilities of passing a chosen structural quality threshold.
Read the full abstract ▾
Combinatorics
The majority of a cactus is prickly
Catarina Carvalho, Mathison Knight and Barnaby Martin
We characterise all (undirected) cacti graphs, and forests of cacti, with an edge-preserving near-unanimity operation, by means of forbidden induced subgraphs. Indeed all such graphs also have an edge-preserving majority operation. We show that said majority can be chosen to be symmetric and explicitly construct it for all forests of cacti without the forbidden induced subgraphs.
Read the full abstract ▾
Systems and Control
A Deficiency-Based Approach for Unbalanced Three-Phase Distribution System Expansion Planning
Eli Brock, Miguel Heleno, Javad Lavaei et al. (+1 more)
Distribution planners typically characterize grid needs through power flow studies and fit upgrades to those needs by hand. More principled one-shot optimization methods either do not scale or misrepresent unbalanced three-phase networks. This paper proposes an iterative planning process that alternates between two problem classes with mature off-the-shelf solvers. An optimal power flow (OPF) problem quantifies the deficiency of a candidate portfolio by optimizing the dispatch of controllable devices and measuring thermal and voltage violations on a common scale, with an optimal value of zero exactly at adequate portfolios. The OPF multipliers give a first-order approximation of the relief each investment candidate would provide. We bound the error of this approximation and show that structural properties of the deficiency guard it against overestimation. Building on this property, a heuristic cutting-plane search uses a mixed-integer linear program to propose each next portfolio from these approximations. On six unbalanced three-phase feeders from the SMART-DS
dataset, the search outperforms a linearized optimization-based
benchmark, and the results agree across five variants of the deficiency.
Read the full abstract ▾
All authors Eli Brock, Miguel Heleno, Javad Lavaei, Somayeh Sojoudi
Optimization and Control
A Composite Representation Theorem for Solution Mappings in Regularized Composite Optimization with Applications to Hausdorff Lipschitz Stability
Peixuan Yang, Wei Yao and Jin Zhang
We study the geometry and stability of solution mappings in regularized composite optimization, allowing multiple minimizers. We establish a composite representation that separates optimal linear images from their corresponding primal solutions through an image mapping and a generalized basis pursuit mapping. Under a fiber-attainment condition, the image mapping admits a reduced optimization formulation; neither representation requires convexity. For convex regularizers, suitable qualification conditions yield a Fenchel-conjugate representation of the reduced value function and a
geometric characterization of the generalized basis pursuit mapping through affine fibers and a fixed lifting set. Using these representations, we establish sufficient conditions for local Hausdorff Lipschitz continuity of possibly set-valued solution mappings. We further show that the primal-space sufficient conditions imply the corresponding image-space conditions, with an additional injectivity requirement arising in the former. The results cover $\ell_1$-type and structured piecewise linear-quadratic regularizers under suitable assumptions. For classical $\ell_1$ regularization, our approach recovers an existing stability result without a finite sign-pattern decomposition. Numerical examples illustrate the recovery of complete solution sets, including a case with a nonpolyhedral graph of the generalized basis pursuit mapping.
Read the full abstract ▾
Econometrics
Dynamic Panel Multinomial Logit Models
Kevin Dano, Bo E. Honoré and Martin Weidner
This paper provides moment conditions for dynamic multinomial logit panel data models with individual specific fixed effects. The moment conditions do not depend on the fixed effects and thus form the basis for generalized method of moments estimation. For a leading case, we provide sufficient conditions under which the moment conditions identify the common parameters, and we illustrate their use in an empirical application to household soft drink purchases. The construction of moment conditions is based on transformed models obtained by permuting outcomes across time periods. For any cyclic permutation, the ratio of the transformed model probability to the true model probability is free of the fixed effects, and comparing two such ratios yields restrictions on the common parameters of interest. This construction is not specific to the multinomial logit. It applies to any dynamic panel data model whose static counterpart admits a symmetric sufficient statistic for the fixed effects. This unifies our main result with analogous results for other models.
Read the full abstract ▾
Analysis of PDEs
Local Turnpike Stability for Two-Population Second-Order Mean-Field Game Systems
Mohamed Ghattassi, Fanze Kong and Nader Masmoudi
This paper establishes a local turnpike property for a two-population, second-order mean-field game system with a quadratic
Hamiltonian, in which the running cost for each population may depend on both population densities. In the case of two populations, imposing a scalar stability condition separately on each species is not sufficient because it neglects cross-interaction effects. The appropriate requirement is instead a
matrix-valued coercivity condition on vector perturbations $(\mu_1,\mu_2)$, which involves the symmetric part of the linearized interaction
matrix together with weighted Poincaré inequalities associated with the two stationary measures. Consequently, in the multi-population framework,
diffusion still acts independently on each population, but the interactions are controlled by the full coupling
matrix. A naive scalar energy functional fails to capture mixed second-variation terms arising from costs that depend on both densities. We prove that
matrix coercivity is precisely equivalent to the strict positivity of this second variation, and we provide a spectral characterization in terms of the principal spectral value of the coupled stationary linearized operator. For linear models with cross-interactions, this yields explicit algebraic conditions that separate stable parameter ranges from unstable ones. In the non-potential framework, the skew-symmetric part of the interaction
matrix leaves the coercivity threshold unchanged, but it can induce transient oscillatory dynamics. Moreover, the scalar turnpike theory extends in a natural, though nontrivial, manner to multi-population systems, where stability is governed by the positivity of a coupled
matrix-valued second variation that encodes both self-interactions and cross-interactions.
Read the full abstract ▾
Soft Condensed Matter
Skyrmion gears: translation-to-rotation conversion of skyrmion clusters along periodic substrates
N. P. Vizarim, C. J. O. Reichhardt and C. Reichhardt
We investigate how the translational motion of a compact skyrmion cluster can be converted into collective rotation through interactions with a periodic obstacle substrate. Using an effective particle model, we consider a seven-skyrmion cluster driven along a one-dimensional array of repulsive obstacle sites. For suitable obstacle spacings and interaction strengths, repeated interactions with the substrate produce coherent clockwise or counterclockwise rotation, giving rise to gear-like cluster dynamics. We characterize these states using the translational velocity,
Hall angle, angular velocity, structural deformation, and a dimensionless gear ratio that relates cluster rotation to motion over the substrate periodicity. Dynamical phase diagrams identify rotating, nonrotating, wall-guided, transverse, and deformed regimes. We further show that coherent rotation persists in the overdamped limit, demonstrating that the conversion mechanism does not require an intrinsic Magnus force and may therefore extend to other driven cluster systems, including colloidal assemblies. These results provide a simple route for controlling collective rotation through structured substrates.
Read the full abstract ▾
Optimization and Control
A Feasible Direction Method with Active-Set Identification for Derivative-Free Optimization
Jeffrey Larson, Matt Menickelly and Lara Zebiane
We propose a model-based trust region method for derivative-free optimization with simulation-based inequality constraints. We are motivated by settings of simulation optimization in which both function evaluations at infeasible points may be unreliable and constraint evaluations are expensive. Our method maintains feasible incumbents and computes steps from a subproblem suitable for a feasible-direction method in order to encourage generation of feasible trial points. To reduce the computational cost of modeling the constraints, we describe an active-set estimation scheme that permits modeling with only a subset of constraints in each subproblem, as opposed to maintaining models of all constraints. We show that every limit point of the incumbents generated by our method is a Fritz-John point, and is additionally a KKT point under the Mangasarian-Fromovitz constraint qualification. Under standard nondegeneracy conditions, the active-set estimate identifies, within finite time, all constraints active at the limit point of the incumbents. Experiments on CUTEst problems show that our method generates feasible trial points far more often than a comparable SQP-based approach and solves more problems within a given budget with fewer function evaluations.
Read the full abstract ▾
Combinatorics
Block Sensitivity and Certificate Complexity of Rotation Symmetric Boolean Functions
Yuan Li and Jing Zhang
We study block sensitivity and certificate complexity of rotation symmetric Boolean functions, with particular emphasis on monomial rotation symmetric Boolean functions. We first show that every odd-degree monomial rotation symmetric Boolean function on $n$ variables satisfies $s=bs=C=n$. We then determine sensitivity, block sensitivity, and certificate complexity for every quadratic monomial rotation symmetric Boolean function. For the quadratic function $Q_{n,k}$ generated by $x_0x_k$, we prove, outside the short-orbit case, that \[ C(Q_{n,k})= \begin{cases} n,&4\mid n/\gcd(n,k),\\ n-\gcd(n,k),&4\nmid n/\gcd(n,k). \end{cases} \] Moreover, \(bs(Q_{n,k})=C(Q_{n,k})\) for every quadratic MRS function, and outside the short-orbit case, strict separation \(s<bs=C\) occurs exactly when \[ \frac{n}{\gcd(n,k)}\equiv2\pmod4. \] Finally, for the consecutive quartic family \[ R_n(x)=\bigoplus_{i\in\mathbb Z_n}x_ix_{i+1}x_{i+2}x_{i+3}, \] we determine all three measures in every residue class modulo $6$, using the sensitivity formulas and, in one residue class, finite-state results established in our companion preprint~\cite{LiZhangRSBFSensitivity}. In particular, for $r=2,3$ and $6q+r\ge5$ we obtain strict separation, \[ \boxed{s(R_{6q+r})=6q<bs(R_{6q+r})=C(R_{6q+r})=6q+1,\qquad r=2,3.} \] Thus this family gives an infinite collection of rotation symmetric Boolean functions for which sensitivity is strictly smaller than block sensitivity, while block sensitivity and certificate complexity coincide.
Read the full abstract ▾
Hardware Architecture
LocalTrain: Integer-Only Local Learning for Energy-Efficient CNN Training on Edge FPGAs
Shiqi Ding, Xuanyu Liang, Xinyang Liu et al. (+4 more)
Local learning reduces cross-layer training dependencies, but an efficient accelerator must still retain operands, accumulate gradients, and select quantization scales. We present an FPGA system that trains a CNN and its regional auxiliary classifiers using a shared integer
matrix engine. Compact errors are expanded on chip, gradients are folded tile by tile, and historical scales permit direct output conversion. At 200 MHz, the KV260 prototype achieves 76.57% accuracy on CIFAR-10 and 91.85% on SVHN at approximately 4.20 W mean module power. Effective
matrix throughput is 31.22 and 31.57 GMAC/s, respectively. Against the same integer method on Jetson Orin Nano, measured training energy is 40.4% and 42.1% lower, with $1.86\times$ and $1.82\times$ longer runtime; measurements use the respective platform power domains. Selected state checks and matching complete-run prediction counts support functional agreement with the software reference.
Read the full abstract ▾
All authors Shiqi Ding, Xuanyu Liang, Xinyang Liu, Boyang Li, Zhiqiang Que, Guosheng Hu, Jiayang Li
Computer Vision
Contrastive learning for video reuse detection
Tristan J.B. Cann and Arjun Biswas
The incorporation of footage from other sources has long been a concern from a copyright perspective, as such usage can harm the interests of rights holders. Recent trends in the “remixing” of social media content and the development of bespoke stock libraries for specialist causes mean that methods to identify content reuse will have a range of applications. In this paper, we develop a framework based on a Siamese network fine tuned using contrastive learning to identify potential sources for key frames in a given
video. This framework makes two contributions to analysis of large
video datasets. Firstly, we demonstrate its utility at reducing complex
video data to a set of key visuals typically less than 10% of the size of the full set. Secondly, we show that these reduced sets allow for efficient and effective identification of component footage. We test this method over a set of
videos compiled from a stock footage library and demonstrate its effectiveness for constituent
video identification with average accuracy greater than 0.95.
Read the full abstract ▾
Cryptography and Security
SEAL-DT: An End-to-End Security Engineering and Assurance Framework for Cross-Domain Digital Twins
Iraklis Varlamis, Stelios Zindros, Christos Sardianos et al. (+3 more)
Cyber-physical systems connect physical processes, software, networks and organizations. Securing them requires a clear account of what happened, what was observed, which actions were authorized and what evidence supports each decision. Digital Twin (DT) descriptions often fail to explain how these elements stay consistent, blur the distinction between physical events and uncertain observations, or detach simulation results from their requirements and assumptions. This paper presents SEAL-DT (Security Engineering and Assurance Lifecycle for Digital Twins), a structured approach to building and evaluating secure DTs. It brings entity definitions, process models, deterministic simulations, data contracts, evidence records, automated tests and reproducible results into one connected model. The main contribution is an end-to-end DT security framework, spanning scenario definition, entity/process modelling, simulation, security decision logic, evidence handling, validation and reproducible reporting. Security decisions and simulation results thus trace back to their data, assumptions and requirements. A DT is more than a copy of system state: it states what it can observe, how accurate and current its information is, where it came from and which decisions it may support. We demonstrate SEAL-DT in five scenarios: Car2X protection, passive IoT-device detection, an industrial control surrogate, lawful acquisition in a smart home, and medical-IoT monitoring that preserves continuity of care. Each includes formal models, deterministic and negative experiments, uncertainty analyses, schema-valid fixtures, technical reports and verified manifests, testing multiple configurations under normal and adverse conditions. Results show that the framework supports consistent model construction, traceable execution and comparable internal validation, and that security findings link to observable evidence.
Read the full abstract ▾
All authors Iraklis Varlamis, Stelios Zindros, Christos Sardianos, Thomas Lagkas, Panagiotis Sarigiannidis, Georgios Th. Papadopoulos
Machine Learning
Towards Universal Calibration for In-Context Classification
Awik Dhar, Sunita Sarawagi and Arpit Agarwal
In-Context Learning (ICL) has received a lot of interest recently due to its ability to improve classification accuracy for new tasks without having to fine-tune the model over these tasks. However, the impact of ICL on calibration has been far less studied. Recent work has shown that ICL can have a negative impact on calibration due to biased selection of in-context examples. The existing methods for improving ICL calibration require additional task and model-specific training, detracting from the no-fine-tuning convenience of ICL. In this work, we introduce CALIBFORMER, a universal ICL calibrator for classification tasks, that generalizes to new models and tasks without further fine-tuning on these tasks. Specifically, we design a
transformer-based calibrator which uses the model predictions on in-context examples to predict a confidence score for the test input. We show that the calibrator trained on three models and ten tasks, is able to generalize/calibrate ICL predictions on several new tasks and four LLMs unseen during training. Even for frontier models like Gemini, CALIBFORMER is able to obtain over 50% reduction in calibration error.
Read the full abstract ▾
Materials Science
Spin-orbit torque effects in all-electrical spin-wave spectroscopy experiments
José Solano Córdova, Jone Mencos, Richard Schlitz et al. (+10 more)
Electrical excitation and detection of propagating
spin waves are key elements of magnonic devices. While conventional
antennas consisting of non-
magnetic metals such as Au or Cu excite
spin waves predominantly through
radio-frequency (RF) Oersted fields,
antennas comprising Pt can additionally generate
spin-orbit torques (SOTs) via the
spin Hall effect. Recent theoretical work has proposed RF-SOT as an alternative mechanism for coherent
spin-wave excitation, predicting distinct characteristics compared with conventional inductive excitation. Here, we investigate Pt
antennas for the excitation of coherent
spin waves in 105-nm-thick epitaxial $\text{Y}_3\text{Fe}_5\text{O}_{12}$ (YIG) films through two potential mechanisms: an oscillating Oersted field and an oscillating
spin current. The raw experimental transmission spectra are strongly affected by fabrication-induced modifications to the YIG film associated with the etching process together with spatially inhomogeneous interface conditions induced by Pt. After removing these contributions, we find the remaining two wave propagation consistent with a combined excitation from Oersted fields and SOTs.
Read the full abstract ▾
All authors José Solano Córdova, Jone Mencos, Richard Schlitz, Quentin Rossi, Denise Reustlen, Javed Ali, Sebastian Sailler, Michaela Lammel, Matthieu Bailleul, Daniel Stoeffler, Felix Casanova, Sebastian T. B. Goennenwein, Helena Reichlova
Human-Computer Interaction
Particle Selection in Virtual Reality
Boaz Keren-Gil, James Gain, Lucia Marchetti et al. (+2 more)
Point-cloud representations of 3D data are used in many fields, from astronomical surveys of
galaxies to medical imaging and
laser scans of buildings. A task common to all of these is interactive selection: picking out a meaningful subset of points for closer analysis. The data is inherently 3D, yet the traditional workflow uses a 2D mouse on a 2D projection and inherits the limitations of a flat screen. This paper compares Virtual Reality (VR) and desktop environments for 3D particle selection, asking whether VR's six-degree-of-freedom hand interaction yields measurable gains in accuracy, speed, usability, workload and flow. In a counterbalanced within-subjects experiment (N = 34), participants selected 90 target
galaxies from a catalogue of 16,816
galaxies, using mouse-driven selection with keyboard camera control on the desktop and a hand-held spherical brush that adds or removes particles in VR. We measured task accuracy, completion time, and three standard questionnaires: the System Usability Scale (SUS), the NASA Task Load Index (NASA-TLX) and the Flow State Scale (FSS). Accuracy was comparable (91.6% VR vs. 93.1% desktop; p = 0.858), but every other measure favoured VR. Tasks were 23% faster (p = 0.001, d = 0.93), usability was higher (SUS 84.6 vs. 54.0; d = -1.78), workload was lower (NASA-TLX 20.4 vs. 34.7; d = 0.90), and flow was higher (FSS 83.2 vs. 70.2; d = -1.23). Because the underlying problem, making a 3D selection through a 2D interface, is not specific to astronomy, we expect the pattern to carry over to point-cloud selection in other fields.
Read the full abstract ▾
All authors Boaz Keren-Gil, James Gain, Lucia Marchetti, Patrick Marais, Alexander Sivitilli
Artificial Intelligence
SFT Leaves a Reproducible Signature in Activation Geometry: Calibrating Post-Training Interpretability Against Run-to-Run Noise
Xupeng Chen, Chenqian Le, Jiayi Cheng et al. (+2 more)
A model diff credits post-training with whatever separates a base checkpoint from one post-trained descendant, presuming a rerun would land close by. On the sparse co-activation graph of Qwen3-1.7B it does not: two LoRA-SFT reruns sit over half as far apart as either sits from the base. We score each base-to-post contrast against the run-to-run band of its own recipe, and show that $k$ runs certify no better than $p=1/(k+1)$ under run exchangeability, however large the effect, so the band $z$-score is a descriptive scale. SFT leaves a large, reproducible signature: base $\to$ SFT displaces the graph by 17.8 pairwise-distance SDs beyond the mean distance between four independent LoRA-SFT runs, and the pre-registered
estimator replicates on Llama-3.1-8B with Llama-Scope. At 29 steps each, four SFT and four DPO runs occupy disjoint displacement ranges, a gap $5.0\times$ and $5.6\times$ the spreads of the two arms, without any band. All bands use graphs built on one GPU architecture. Three other readings stop short of a verdict: DPO band counts (3/4, 1/4, 3/4 contrasts outside at 12, 29, 60 steps) show no consistent dose ordering while its displacement stays flat; light GRPO is a secondary non-detection, not an absence, and we run no equivalence test; and the graph concentration read as an RL signature is already in the zero-RL base. A pre-registered causal sweep at three edge budgets on 4,816 held-out
prompts found the displaced sources load-bearing ($12\times$ a matched random control in mean, $3.5\times$ in median at 10,000 edges), but their stage-specificity interaction reversed at two budgets and the third failed its faithfulness gate, so we claim calibrated structural displacement, not a changed mechanism. Scope: one backbone, one SFT recipe and general-text probe states for the full protocol.
Read the full abstract ▾
All authors Xupeng Chen, Chenqian Le, Jiayi Cheng, Runhao Li, Michael Yue
General Relativity and Quantum Cosmology
Bartnik's splitting conjecture and the cosmological wall: a Cauchy Busemann function approach
E. Minguzzi
Bartnik's
conjecture states that a timelike geodesically complete cosmological spacetime satisfying the timelike convergence condition splits as a static product. We show that any counterexample exhibits a peculiar global causal structure: two disjoint open regions, the future and the past cosmological developments, separated by a non-empty closed causally convex set: the cosmological wall. The developments are bounded by $C^0$ null hypersurfaces, the cosmological horizons, and a component of each horizon contains a lightlike line. These objects are natural and hence carry physical significance, as they do not depend on the choice of Cauchy hypersurface. They can be defined through the Cauchy Busemann functions $b^\pm$ of a compact Cauchy hypersurface $\Sigma$, the developments being their finiteness domains. Using a product trick we show that $b^+$ is the graph function of a Cauchy horizon in $M\times \mathbb{R}$, and so inherits the regularity and dynamical properties of horizons: $b^+$ is continuous, locally Lipschitz and locally semi-concave on its finiteness domain, and its calibrated curves are timelike rays. We prove that if the two cosmological developments intersect, which happens whenever $b^+$ is $C^1$ on its finiteness domain, then the spacetime contains a timelike line and hence splits, a result which recovers several positive results of the literature. We also obtain a localized
rigidity version of Hawking's singularity
theorem: if some compact Cauchy hypersurface has non-positive mean
curvature on the future development, the spacetime splits. Hence, in a counterexample, every compact Cauchy hypersurface expands somewhere in the future development and contracts somewhere in the past one. If the wall has empty interior, it coincides with both horizons and is ruled by lightlike lines, and under null completeness its components are smooth totally geodesic null hypersurfaces.
Read the full abstract ▾
Networking
MCCL: Measurement-based Collective Communication Library
Michele Ferrero, Gabriele Castellano, Paolo Medagliani et al. (+3 more)
Collective communication is a critical performance bottleneck in large-scale distributed training. Existing libraries and synthesis frameworks compute open-loop schedules once, assuming a static
topology and peak link capacities. In production, however, persistent gray failures can silently degrade link throughput and latency for hours, causing pre-computed schedules to lose much of their efficiency during execution. In this paper, we present Measurement-based Collective Communication Library (MCCL), a collective communication library that closes the loop between runtime network measurement and dynamic schedule synthesis. MCCL passively measures GPU-to-GPU throughput, detects persistent degradation, re-synthesizes the affected communication-intensive schedules with warm-started synthesizers, and hot-swaps them without interrupting training. We validate the complete loop on a multi-node hardware prototype and use simulations to evaluate cluster-scale scenarios. In our experiments, MCCL incurs negligible overhead (<2%) while maintaining high measurement accuracy (<1.2% error) and low schedule hot-swap time (about 12 ms). Notably, under gray failures, MCCL reduces training time by up to 4.99x over static schedules on simulated training workloads for state-of-the-art LLMs such as Llama and DeepSeek.
Read the full abstract ▾
All authors Michele Ferrero, Gabriele Castellano, Paolo Medagliani, Sebastien Martin, Roberto Morabito, Massimo Gallo