Computation and Language
NADI 2026: The Second Multidialectal Arabic Speech Processing Shared Task
Peter Sullivan, Bashar Talafha, Ahmed Ashraf et al. (+11 more)
NADI 2026 is the seventh edition of the Nuanced Arabic Dialect Identification (NADI) shared task series and the second dedicated to multidialectal Arabic
speech processing. This edition comprises five tasks and eight subtasks spanning Automatic
Speech Recognition (ASR), Spoken Dialect Identification (SDID), Text-to-
Speech (TTS), Spoken
Language Translation (SLT), and Spoken
Language Understanding (SLU). NADI 2026 emphasizes realistic evaluation through low-bandwidth, mixed-dialect, code-switched, out-of-domain, and zero-shot settings, while introducing TTS, SLT, and SLU to the series for the first time. The shared task attracted 21 participating teams from at least 13 countries, with 48 test-phase submissions and 14 submitted system-description papers. Results show that out-of-domain generalization remains a major bottleneck and highlight the effectiveness of recent Arabic-specialized
speech models,
multimodal dialect identification approaches, and ensemble methods. Overall, NADI 2026 provides a broader and more challenging
benchmark for robust Arabic dialect
speech processing.
Read the full abstract ▾
All authors Peter Sullivan, Bashar Talafha, Ahmed Ashraf, Fethi Bougares, Haroun Elleuch, Chiyu Zhang, AbdelRahim Elmadany, Youssef Mohamed, Salima Mdhaffar, Yannick Estève, Mohamed Elhoseiny, Hamzah Luqman, Nizar Habash, Muhammad Abdul-Mageed
Distributed Computing
Crossflow: Prefill-Decode Elasticity for Agentic LLM Serving
Yi Xu, Ehsan K. Ardestani, Wenyin Fu et al. (+7 more)
As serving capacity demand surpasses that of training, serving efficiency becomes increasingly important. Prefill-decode (P/D) disaggregation improves serving efficiency through specialization and isolation of the two phases. These benefits rest on a static partitioning. Phase demand, however, is not static. We observe that in a large LLM fleet the ratio of uncached input to output
tokens has peak-to-mean ratios up to 4.7x at minute timescales, and that in a public agentic trace the hourly ratio spans a median 24.5x within a single day, while reassigning a replica takes tens of minutes. Agentic traffic sharpens the mismatch. Sizing each pool at its ninety-fifth percentile leaves up to 17% of cluster capacity unused; sizing below it converts the same imbalance into queueing and unrealized throughput. We present Crossflow, which makes this boundary elastic without changing node roles. Each decode node publishes a short-lived, revocable lease that bounds local-prefill compute, KV capacity, transfer work, and projected output. Across public and internal traces, Crossflow improves
token throughput by 16.2-17.4% on
geometric mean over static P/D, and by up to 43.4% at high load, while reducing mean TTFT at every evaluated point.
Read the full abstract ▾
All authors Yi Xu, Ehsan K. Ardestani, Wenyin Fu, Martin Schatz, Krishna Malladi, Zhan Shu, Adnan Aziz, Shobhit Kanaujia, Ajit Mathews, Chunqiang Tang
Number Theory
Galois-Orbit Structure, Ramanujan Sums, and Stable-Range Collapse for Cyclotomic Cosine Formulas
Juan D. Vélez and Carlos A. Cadavid
We study cyclotomic cosine evaluations beyond the fully symmetric setting by combining Galois-orbit structure with finite-frequency expansions. First, for truncation-compatible poly?nomial families over Q, we distinguish invariance of the evaluated value from invariance of the
polynomial formula. Summability by a rational function of the level forces eventual rationality of the evaluated values. Conversely, every rational-valued sequence can be realized by a compatible family whose total degree is uniformly bounded by two, showing that value-level Galois invariance alone carries essentially no stable-range
rigidity. We then introduce explicit orbit-structured pattern sums indexed by multiplicative configurations. For every fixed pattern, at each sufficiently large admissible level n, its cyclotomic cosine evaluation collapses to an affine function n times S0(r,h) minus 2 to the abs(h) power, where S0 is the zero-frequency coefficient of the associated Laurent
polynomial. Consequently, every family uniformly generated from finitely many such patterns by a single
polynomial recipe has eventually
polynomial evaluation. Finally, decomposing the punctured index set into gcd-orbits yields an exact orbitwise formula in terms of Ramanujan sums. The semiprime and
prime-power cases become transparent specializations of this general identity. These results identify both the obstruction to extending symmetric
rigidity from value-level Galois invariance and a concrete non-symmetric class exhibiting uniform stable-range collapse
Read the full abstract ▾
Signal Processing
AI-Enabled Wireless Propagation Modeling and Radio Environment Maps for 5G Aerial Wireless Networks
Gautham Reddy, Kürşat Tekb{\i}y{\i}k, Bryton Petersen et al. (+3 more)
With the gaining prominence of aerial mobility applications, their success depends on the seamless integration of terrestrial and non-terrestrial network connectivity. However, providing reliable connectivity from terrestrial telecommunication networks remains challenging due to multi-
cell interference from base stations (BSs) under line-of-sight (LoS) conditions to unmanned aerial vehicles (
UAVs), coverage holes caused by
antenna sidelobe degradation, localized multipath fading effects, and the high-speed dynamics of aerial users. To model such complexities, often exacerbated by sparse real-world data, this work proposes a dual-stage
radio environment map (REM) framework. Our approach physically decouples the channel modeling, where a spatial
Transformer first anchors the deterministic, large-scale path loss geometry, while a gated recurrent unit (GRU) subsequently extrapolates the
stochastic, localized fast- fading deviations. By reformulating 3D spatial interpolation as a 1D radial sequence prediction task, the framework inherently aligns with the physics of propagation. We evaluate the proposed framework against state-of-the-art baselines, including 3D Kriging, UNet, Mamba, and Inception, using empirical 5G
datasets. The results demonstrate improved intra-site generalization across diverse altitudes, user dynamics, and reference signal received power (RSRP)
datasets, achieving signal-strength predictions with errors near 3 dB and REM spatial-similarity indices exceeding 0.75. Finally, we examine the influence of REMs and channel rank conditions on
UAV channel quality, underscoring the necessity of reliable channel modeling for robust aerial connectivity.
Read the full abstract ▾
All authors Gautham Reddy, Kürşat Tekb{\i}y{\i}k, Bryton Petersen, Antoine Lesage-Landry, Gunes Karabulut Kurt, Ismail Güvenç
Signal Processing
Automated feature-region selection for soft-sensor development from spectral-like measurements
Sebastián Espinel-Ríos and Wenchao Duan
Sustainable production increasingly relies on process analytical chemistry and process analytical technology to support monitoring, control, and automation. Techniques used in these contexts, including Raman spectroscopy, infrared spectroscopy, and electrochemical voltammetry, generate high-dimensional signals ordered along physical measurement axes, referred to here as spectral-like measurements. These signals can contain redundant, weakly informative, and noisy regions, complicating the development of data-driven soft sensors to map them to process variables. Selecting informative regions is nontrivial, as visually prominent regions are not necessarily the most predictive, while synergistic effects among regions cannot be readily inferred. Here, we introduce an automated feature-region selection framework that identifies a parsimonious set of contiguous regions while preserving channel ordering. The framework combines channel-level target correlation and a signal-to-noise indicator into a latent information fingerprint. Variable-width candidate intervals are derived from peaks in this fingerprint, and their combinations are ranked using held-out validation data. Final selection favours models using fewer channels among candidates with comparable predictive performance. The framework is demonstrated using cyclic voltammetric measurements of glucose acquired with a gold electrode sensor across four progressively broader nominal concentration ranges up to 350 g/L. Gaussian process regression is used within the framework to accommodate possible nonlinear relationships, account for observation noise, and quantify epistemic uncertainty. The selected models retained only 11-42 of the original 400 channels and reduced held-out test root mean squared error by 88.3-94.6% relative to full-feature Gaussian process regression
benchmarks.
Read the full abstract ▾
Astrophysics of Galaxies
CECILIA: Multi-line Constraints on Excess Nebular Emission from Low-Ionization Gas at Cosmic Noon
Caroline von Raesfeld, Allison L. Strom, Audrey Clarendon et al. (+7 more)
Spectroscopy from JWST has offered unique insights into elemental abundance patterns at
Cosmic Noon (z~2-3), most notably a growing census of sub-
solar S/O. These new measurements, coupled with new observations of [O I]{\lambda}6302 at z>1, motivate an updated study of neutral and low-ionization emission within
Cosmic Noon
galaxies. In this letter, we present a new analysis of [O I]{\lambda}6302 and [S II]{\lambda}{\lambda}6718,33 in 19
galaxies observed using ultra-deep JWST/NIRSpec spectroscopy as part of the CECILIA survey. We find that it is necessary to consider sub-
solar abundance patterns in sulfur at high-z when analyzing
galaxies on the S2-BPT, and that this correction reveals the potential of additional emission sources in CECILIA
galaxies. Using the photoionization model emulator Cue, we find that ~48% of observed [S II] and ~55% of observed [O I] emission in CECILIA
galaxies cannot be reproduced by conventional HII region models. We find no clear evidence for
turbulence or shocks driving this excess emission when comparing to existing models. Our findings imply that this excess emission may originate in diffuse ionized gas outside of H II regions, though its nature may be different than that of low-z DIG. Further, we find that including emission from gas outside of
star-forming regions can incur a small bias up to 0.1 dex in 12+log(S/H).
Read the full abstract ▾
All authors Caroline von Raesfeld, Allison L. Strom, Audrey Clarendon, Noah S.J. Rogers, Yijia Li, Nathalie A. Korhonen Cuestas, Gwen C. Rudie, Ryan F. Trainor, Menelaos Raptis, Zhuyun Zhuang
Pattern Formation and Solitons
Robustness of mKdV Breathers: A Numerical Perspective
Chandler Haight, Svetlana Roudenko, Diana Son et al. (+1 more)
We systematically investigate breather solutions in the modified Korteweg-de Vries (mKdV) equation via numerical simulations. We show that the breather solutions are stable under a variety of perturbations, including amplitude changes, modifications of internal parameters, and perturbations of the nonlinearity of the equation. Our results are not only consistent with the known analytical studies on the stability of breathers in integrable systems, extending them further, but also provide numerical evidence that such breather-type structures can persist and remain stable over the simulated time scales in some non-integrable settings with symmetric potentials.
Read the full abstract ▾
All authors Chandler Haight, Svetlana Roudenko, Diana Son, Kai Yang
Mesoscale and Nanoscale Physics
Spectroscopic signatures of persistent exciton condensation in a bulk magnetic topological insulator
Paulina Majchrzak, Chakradhar Sahoo, Manuel Tuniz et al. (+7 more)
Exciton condensates are long-sought correlated
quantum states arising from macroscopic coherence of bound electron-hole pairs. Although
equilibrium and transient excitonic states have been reported in several material platforms, direct evidence for a light-induced
exciton condensate state has been challenging to achieve as an intrinsic property of a bulk
quantum material. Here, we use time- and angle-resolved photoemission spectroscopy to investigate long-lived photoexcited carriers in the intrinsic
magnetic topological
insulator MnBi$_2$Te$_4$ with the chemical potential tuned to the topological surface state by Sb substitution. Following optical excitation, we observe the delayed emergence of a transient state whose formation coincides with depopulation of the bulk conduction band and whose lifetime extends to the microsecond timescale. The quasiparticle dispersion exhibits a pronounced flattening and develops a Mexican-hat-like profile. These spectral signatures are consistent with the formation of a metastable excitonic condensate state. Our results establish
magnetic topological
insulators as a promising platform for investigating long-lived photoinduced many-body states and their interplay with
topology and magnetism.
Read the full abstract ▾
All authors Paulina Majchrzak, Chakradhar Sahoo, Manuel Tuniz, Wibke Bronsch, Denny Puntel, Federico Cilento, Xing-Chen Pan, Jakob Kjærulff Svaneborg, Yong P. Chen, Søren Ulstrup
Logic
Fine Selection for Intuitionistic Modal Logic
David Fernández-Duque
We extend Fine's selection method to the setting of intuitionistic modal logic and use it to provide a model-theoretic proof that Fischer Servi-style intuitionistic $\sf K4$ has the finite model property.
Robotics
Fast Direction-Conditioned Reachability for Motion Prediction Under Model Uncertainty
Hrishav Das and Melkior Ornik
To avoid collisions, a
robot must repeatedly predict where nearby
agents may move, usually with an imperfect model of their dynamics. Reachable sets provide such predictions, but computing them when the system
matrices themselves are uncertain can become computationally expensive and conservative for frequent replanning. Moreover, a planner often needs to know only how far an
agent can move in one particular direction, for example toward the
robot, rather than the complete reachable set. We propose a direction-conditioned reachability method for linear systems with uncertain state and input
matrices. Given a query direction $d$, the method selects one admissible model $(A^\star,B^\star)$ whose reachable set extends nearly as far along $d$ as the reachable set of the entire uncertain model family, and then computes the reachable set of only this model with a standard reachability solver. On an uncertain linearized bicycle model, the complete selection-and-computation pipeline is about three times faster than computing the reachable set of the full uncertain family in the CORA toolbox, while its extent along $d$ is within $5\%$ of the full family's in the reported directions. We also use the method in a closed-loop multi-vehicle simulation in which the
robot queries, at each replanning step, how far each nearby vehicle can move toward it, and replans to avoid the resulting sets.
Read the full abstract ▾
Computer Vision
Pro-Bench: Prompt-Robust Open-Vocabulary Visual Grounding Across Real-World Heterogeneous Environments
Linus Nwankwo, Muslim Alaran, Christian Rauch et al. (+2 more)
Open-vocabulary visual grounding enables
robots to localise task-relevant entities from natural-
language queries without dependence on predefined perceptual taxonomies. However, existing
benchmarks largely rely on short category labels and web-scraped imagery, leaving it unclear whether open-vocabulary models can robustly ground diverse queries and visual conditions under real deployments. We introduce Pro-Bench, a
prompt-conditioned
benchmark for open-vocabulary visual grounding in heterogeneous, real-world environments. Pro-Bench includes $13k+$ RGB frames from independent robotic domains (subterranean, industrial, indoor, outdoor, urban), with $74.5k$ manual instance annotations and $515$ target queries covering categorical, attributive, relational, affordance, state, part-whole, negative, and compositional semantics. We benchmarked $16$ open-vocabulary model configurations in strict zero-shot inference, measuring localisation accuracy across IoU thresholds, end-to-end inference latency,
prompt-induced performance variation, and target recovery consistency. Our results show that
prompt-robustness is strongly architecture-dependent. Most model configurations ($10/16$) perform best with short category labels, whereas free-form queries yield the highest accuracy for only one. Moreover, similar aggregate mAP can conceal substantial differences in consistent target recovery across reformulations. Pro-Bench enables systematic evaluation of these gaps and supports
prompt-robust visual grounding. Pro-Bench: https://pro-bench.github.io/.
Read the full abstract ▾
All authors Linus Nwankwo, Muslim Alaran, Christian Rauch, Stanley Chukwuebuka Obilikpa, Elmar Rueckert
Statistical Methodology
Formulating Cross-World Mediation Estimands Through Single-World Mixtures
Razieh Nabi and David Benkeser
In causal mediation analysis, natural direct and indirect effects are defined through nested counterfactuals that combine the outcome under one exposure level with the mediator value under another and are therefore inherently cross-world. Their canonical identification additionally relies on cross-world independence assumptions. Consequently, both the estimands and their identifying assumptions remain controversial. In this paper, we ask what additional single-world structure would be required to re-express cross-world estimands as single-world quantities, identifiable under single-world assumptions. We show that these quantities can be written as mixtures of controlled single-world effects under assumptions involving a susceptibility marker, a possibly latent baseline variable that encodes the "would-be" mediator value under a reference treatment arm. If observed, the marker would make several implications of the marker restrictions empirically testable. These results provide a transparent single-world formulation of cross-world mediation effects while making explicit the assumptions required. Although these conditions clarify how cross-world estimands can be interpreted within a single-world framework, their practical relevance depends on whether such susceptibility markers can be justified in specific applications. We discuss implications for principal stratification and show in the Supplementary Material how the formulation extends to ordered mediators and settings with exposure-induced mediator-outcome confounding.
Read the full abstract ▾
Computers and Society
Quantifying the Occult: A Comparative Study of Hindu and Buddhist Deities Using Machine Learning Methods
Ankit Bhattacharjee
This study introduces a dual-
matrix computational architecture to mathematically quantify the morphological and theological divergence of 196 Hindu and Vajrayana Buddhist esoteric deities. Physical morphology is evaluated via a discrete Gower distance
matrix enhanced by a novel "Cardinality Weighting"
algorithm, while theological function is mapped via dense vector
embeddings generated from Large
Language Model (LLM) semantic expansions, explicitly utilized as a synthetic proxy to mitigate circular
reasoning. The multi-modal topological projections provide algorithmic validation of "iconographic camouflage", demonstrating how distinct visual forms structurally obscure shared cross-tradition functions. Furthermore, I computationally model the "Atin Effect" - serving simultaneously as a psychological observation of sequential cognitive bias and a machine learning
benchmark - demonstrating how high-cardinality esoteric anchors (e.g., a veena or a severed head) override systemic theological disparities to mathematically cluster orthodox and Tantric entities. Cross-tradition spatial analysis establishes that the highest esoteric manifestations, such as the Hindu Chinnamasta and the Buddhist Chinnamunda, share a near-identical mathematical coordinate across both visual ($D_G = 0.288$) and semantic ($D_C = 0.068$) boundaries, indicating a 1:1 esoteric transfer. By open-sourcing this architecture, I provide a scalable, unsupervised machine learning tool for Digital Humanities scholars and comparative theologians to rigorously map latent structural continuities across qualitative cultural corpora.
Read the full abstract ▾
High Energy Physics (Phenomenology)
Inverse phase transitions via dark sector chemical equilibration
Esau Cervantes, Felix Kahlhoefer, Jonas Matuszak et al. (+1 more)
We study the cosmological
evolution of the effective potential in a dark sector with a spontaneously broken Abelian gauge
symmetry, populated out of
equilibrium through the freeze-in mechanism. Initially dilute and in the broken phase, the dark sector is described by a temperature as well as a chemical potential that reflects the low occupancy of dark sector states. We show that the thermal corrections to the effective potential depend on both of these quantities, and that they are suppressed while the sector remains dilute, but grow as internal number-changing reactions increase the occupancy and drive the system toward chemical
equilibrium. We solve the coupled
evolution of the temperature and chemical potential and the resulting impact on the effective potential to show that, for sufficiently large portal couplings, chemical equilibration can evolve the effective potential across a first-order barrier and drive the dark sector toward its symmetric phase through an inverse phase transition.
Read the full abstract ▾
All authors Esau Cervantes, Felix Kahlhoefer, Jonas Matuszak, Santiago Rosellon
Quantum Physics
Programmable Coherent Memory Kernels for Quantum Reservoir Computing in Waveguide QED
Hong Jiang, Yu Wu, Yue Ban et al. (+2 more)
Reservoir computing requires not only long
memory, but also
memory distributed over the timescales relevant to a given task. In most
quantum reservoir computing architectures, however, the
memory profile emerges from intrinsic dynamics and cannot be directly engineered. Here we introduce a waveguide-QED
quantum reservoir in which coherent
memory return is controlled by the coupling geometry. The reservoir consists of a driven Kerr-nonlinear resonator coupled to a one-dimensional waveguide at spatially separated points. An unmeasured propagating field carries information about past inputs and coherently returns it to the resonator after a finite propagation time, forming a non-Markovian feedback channel without intermediate measurement or classical reinjection. Delay-resolved
memory benchmarks reveal two complementary functions: waveguide propagation determines when past information returns, whereas Kerr dynamics convert the returned field into nonlinear temporal features accessible through a linear readout. NARMA-$n$ prediction shows that the optimal feedback delay shifts systematically with task order and approximately tracks the explicit input-product lag. These results establish coherent delayed feedback as a mechanism for engineering task-relevant
memory in
quantum reservoir computing.
Read the full abstract ▾
All authors Hong Jiang, Yu Wu, Yue Ban, Xi Chen, Juan José García-Ripoll
Dynamical Systems
Sharp effective equidistribution of closed geodesics in compact hyperbolic surfaces
Junehyuk Jung, Insung Park and Peter Zenz
We prove a sharp effective equidistribution
theorem for closed geodesics on a compact
hyperbolic surface. This is achieved by carefully analyzing the Selberg trace formula twisted by a Maass eigenfunction, which we refer to as Zelditch's trace formula, following Zelditch's observation that the
geometric side can be expressed as a sum of period integrals of eigenfunctions on closed geodesics. The main technical input is a sharp upper bound for the
matrix coefficients $\langle \phi, \varphi^2 \rangle$ uniformly in $t_\phi\to \infty$ and $t_\varphi \to \infty$, where $t_\phi$ and $t_\varphi$ are the eigenparameters corresponding to Maass forms $\phi$ and $\varphi$ on $X$.
Read the full abstract ▾
Robotics
The Gaussian Is Enough: Flow-Matching Priors Do Not Help When Fine-Tuning Large Behavior Models
Chen Xu, Rishi Shah, Hadas Kress-Gazit et al. (+2 more)
Modern
robot imitation learning increasingly relies on generative policies based on
diffusion or flow-matching models, which generate actions by transforming samples from a prior distribution. A key question is whether the choice of prior matters. Replacing the standard Gaussian with a closer-to-target, non-Gaussian prior has been shown to substantially improve performance when training from scratch. A natural next step is to ask whether these gains transfer to fine-tuning pretrained Large Behavior Models (LBMs) such as LBM 1.0, $\pi_{0.5}$, and GR00T~N1.5, where one might expect even larger gains. Surprisingly, we find that this is not the case, except possibly at very low fine-tuning data fractions. Across over 100K simulation rollouts spanning all three aforementioned LBMs on 40+ tasks in two simulation platforms, and 1250 hardware rollouts on five bimanual manipulation tasks, non-Gaussian priors that are demonstrably closer to the target yield statistically indistinguishable or worse fine-tuning performance than a standard Gaussian prior. Diagnostic analyses suggest why: fine-tuned imitation learning policies converge to similar action predictions across priors, despite their fine-tuned encoder
embeddings diverging substantially from the pretrained
embeddings and each other. A learning-rate ablation further confirms that encoder training is the dominant factor in fine-tuning performance, substantially outweighing the effect of prior choice. We conclude with concrete directions for future research on when and why learned priors might still matter in fine-tuning. Project page: https://cxu-tri.github.io/non_gaussian_FT/
Read the full abstract ▾
All authors Chen Xu, Rishi Shah, Hadas Kress-Gazit, Haruki Nishimura, Masha Itkina
Software Engineering
Does Graph Structure Earn Its Place in Microservice Root-Cause Analysis? A Controlled Study on RCAEval, and What the Benchmark Was Really Measuring
Imad Buljić
Graph neural networks dominate recent work on microservice root-cause analysis, yet recent results question whether the graph contributes. Those results compare whole pipelines, so when a flat model wins one cannot tell whether structure is useless or redundant. We run the comparison they imply on RCAEval: three learned arms with identical features, optimiser, validation split, early-stopping rule and scoring head, in which a single term separates the graph arms. Across two RCAEval
benchmarks, two
topology sources and four regimes we find no reliable graph-specific effect: in-distribution the graph model leads the conventional flat model by 0.003 Avg@5 (p = 0.844, n = 6 disjoint folds).
Auditing the pipeline surfaced two
benchmark properties that condition any result on it. RCAEval injects faults into only five services per system while exposing 12 to 70 in telemetry, and the headline metric is Avg@5: a ranker reading no telemetry at all places the true culprit in the top five on 99.7 percent of held-out incidents, scoring Avg@5 0.488. That prior, not the uniform-random 0.137, is the honest in-distribution floor, and it collapses to 0.192 across systems. The second property is a non-uniform column schema that silently zeroes telemetry for most RE1 cases. We reproduce a published baseline, BARO, RCAEval's own reference implementation; on the one system with a clean schema it reaches similar aggregate accuracy to our heuristic, within 0.004, under a different scoring rule. The audit motivated a new model. PSC-GRCA separates a candidate score into a system prior, telemetry evidence and a centred graph residual, and reaches mean Avg@5 0.915 against 0.864 for the flat baseline, while its ablations locate most of the gain in the prior term rather than the graph. We close with a twelve-item checklist for graph-versus-flat ablation studies, distilled from sixty-two recorded defects.
Read the full abstract ▾
Robotics
HiRE: Hindsight Reward Editing for Policy Finetuning
Haoyi Niu, Zhengtao Han, Yufeng Ji et al. (+2 more)
Pre-trained
robot policies always require finetuning to adapt to specific environments. Reinforcement Learning (RL) offers high performance potential because it improves action optimality rather than simply mimicking data. However, such potential depends heavily on reward quality. Sparse rewards lack process feedback, human-designed rewards are costly and biased, and semantic rewards from foundation representations are often not control-centric. We propose Hindsight Reward Editing (HiRE), a training-free framework to break this reward bottleneck. HiRE bridges the broad knowledge of foundation representation models with physical control awareness, by contrasting successful and failed trajectories in hindsight. It calibrates foundation representation models by identifying "trap states" that are predicted as high-rewarding states yet eventually result in failure, and vice versa. HiRE explicitly penalizes these traps while boosting rewards for critical successful states. This approach can be flexibly compatible with any foundation representations and RL
algorithms. Experiments show that HiRE consistently outperforms other reward recipes by delivering dense, control-aware feedback that prevents value function collapse and reward hacking, thereby achieving superior sample efficiency, stable policy updates, and higher performance ceilings, e.g., at least 3x performance of the base policies. Qualitative results are at https://hire-project.github.io .
Read the full abstract ▾
All authors Haoyi Niu, Zhengtao Han, Yufeng Ji, Zhongyu Li, Koushil Sreenath
Machine Learning
ChipMEM: Verification-Grounded Memory for EDA Agents
Abdulrahman AlRabah, Joshua Mabry, Dilek Hakkani-Tür et al. (+4 more)
Large
language model (LLM)-based
agents use Electronic Design Automation (EDA) tools to generate and revise register-transfer-level (RTL) designs under synthesis and verification feedback. Recent methods learn from this feedback by distilling reusable skills from execution traces or by training on rewards derived from EDA-tools. Both methods are typically evaluated on the tasks that produced the experience. Repeated access to
benchmark feedback on the same task can reward task-specific revision rather than creating reusable knowledge that transfers. We introduce ChipMEM, a verification-grounded
memory layer for EDA
agents. It combines cross-task procedural
memory with within-trajectory statistical guidance. Its procedural component distills and stores a skill only after it passes synthesis, simulation, or formal checks, rather than relying on model self-assessments. A
Bayesian component maintains hierarchical Beta estimates over tool-call outcomes and ranks recovery strategies that succeeded under comparable errors. A common adapter applies the same
memory interface to RTL optimization and testbench-generation
agents while preserving each domain's tools and acceptance criteria. We measure performance on training tasks and evaluate whether learned skills transfer to unseen tasks. On RTLRewriter-Bench, under matched model and tool settings, ChipMEM produces equivalence-passing outputs on 39/54 scored designs versus 35/54 without
memory; on the 49-design short suite, mean area improvement is 8.69% versus 5.66%. On held-out CVDP tasks, ChipMEM with a frozen procedural library achieves 20/20 accepted outcomes versus 18/20 without
memory in a single evaluation per setting.
Read the full abstract ▾
All authors Abdulrahman AlRabah, Joshua Mabry, Dilek Hakkani-Tür, Abdussalam Alawini, Hamid Shojaei, Kartik Hegde, Sandesh Adhikary
Systems and Control
Evaluating Grid Strength with Rising Penetration of Inverter Based Resources
Mohammed N. Khamees, Gad Ilunga, Zhongxia Zhang et al. (+1 more)
Over the years, Dominion Energy Virginia (DEV) has experienced recurring power quality disturbances, including voltage and power oscillations, particularly in areas with significant inverter-based generation. While these phenomena suggest a potential relationship to system strength, no definitive correlation has yet been established. This paper presents methodologies for assessing system strength across the DEV network and identifying regions that may be vulnerable to such disturbances. To this end, three grid-strength metrics are evaluated: the Simple Short-
Circuit Ratio (SSCR), the Weighted Short-
Circuit Ratio (WSCR), and the Composite Short-
Circuit Ratio (CSCR). Historical event data will be analyzed using steady-state snapshots from DEV's proprietary Analysis on Demand (ANODE) platform, which provides 10-minute interval load-flow data. These snapshots will be processed using a developed software tool to compute system strength metrics. The analysis will investigate potential correlations between system strength and observed disturbances. Where such correlations are identified, the study will further evaluate the impact of deploying synchronous condensers (SynCons) as a mitigation strategy.
Read the full abstract ▾
All authors Mohammed N. Khamees, Gad Ilunga, Zhongxia Zhang, Richard E. Tuck
Mesoscale and Nanoscale Physics
Thermodynamic Uncertainty Relations in Chaotic Andreev Billiards
I. R. A. C. Lucena, T. J. A. Mori, M. M. Soares et al. (+2 more)
We investigate how particle-hole
symmetry,
quantum interference, and tunnel barriers shape thermodynamic uncertainty relations in chaotic Andreev billiards. Using random-
matrix theory and the Mahaux-Weidenmüller scattering approach, we study charge conductance and shot noise across the four Altland-Zirnbauer
symmetry classes, from the single-channel extreme
quantum limit to the multichannel semiclassical regime and from ideal to opaque contacts. We characterize thermodynamic precision through two complementary ensemble observables: either by averaging the sample-resolved noise-to-conductance ratio, or from the separately averaging noise and conductance. Their pronounced discrepancy in the extreme
quantum regime reveals the non-self-averaging character of mesoscopic transport and persists over a broad range of barrier transparencies. For ideal contacts, the first provides a sensitive fingerprint of the Altland-Zirnbauer
symmetry class. In the opaque regime, this hierarchy changes. Despite these strong
symmetry- and barrier-dependent effects, the standard thermodynamic uncertainty relation remains satisfied throughout all regimes investigated. Our results establish thermodynamic uncertainty as a
symmetry-sensitive probe of universal transport in chaotic normal-superconducting systems.
Read the full abstract ▾
All authors I. R. A. C. Lucena, T. J. A. Mori, M. M. Soares, D. Bazeia, A. R. Rocha
Computation and Language
What Changes When Fact-Verification Scores Improve? Evidence and Answer Accounting Across Trained Verifiers and LLMs
Han Chen and Yingrui Li
A joint fact-verification score assesses answers and submitted evidence together. When the score improves, how much of the gain remains if the answers are held fixed? On FEVEROUS, strict score is the percentage of claims with a correct answer and a complete annotated evidence group in the submitted evidence. Across four trained DeBERTa checkpoints and 7,890 claims, replacing DCUF evidence with UnifEE evidence raises strict score by 9.61 percentage points, compared with 1.96 percentage points in answer accuracy. The paired 95% interval for the strict-score gain is [8.77, 10.43], conditional on these checkpoints. Replacing only the evidence passed to the scorer accounts for 7.92 or 9.08 percentage points when we retain the answers generated from DCUF or UnifEE evidence, respectively. To examine how this evidence gain depends on evaluation choices, we generate 470,400 responses from two 8B LLMs on FEVER, FEVEROUS, and SciFact under two answer formats and two context budgets. Increasing context from 256 to 2,048
tokens raises the fixed-answer evidence gain on FEVEROUS by 3.84 and 3.10 percentage points for Qwen and Llama, respectively. The effects fall short of the prespecified cross-
dataset criterion, while some intervals extend beyond the two-point small-effect bound. Post-hoc analyses quantify changes in answers and submitted evidence, and show when aggregate accuracy and evidence-coverage rates miss the claim-level pattern. The four answer-evidence score combinations reveal changes that endpoint and aggregate metrics leave unresolved.
Read the full abstract ▾
Computers and Society
Student Use of LLMs and the Limits of AI-Generated Question Difficulty in Data Science Courses
Yuan An and Lei Wang
This paper presents a multi-source classroom study conducted during a 10-week quarter in data science courses at Drexel University. We first investigate the behaviors of student engagement with large
language models (LLMs) using four surveys across three data science courses. Second, we evaluate the construct validity of multiple-choice questions (MCQs) generated by an LLM for in-lecture
retrieval practice. Based on 378 authored questions (311 deployed, producing 7{,}888 student responses), we analyze whether the difficulty ratings assigned by an LLM match empirical item difficulty. Our study shows that student engagement with LLMs varied across courses and increased over the term. Although students expressed high satisfaction and reported saving considerable time, their perception of deep learning benefits declined, and many noted a tendency toward over-reliance. Regarding the difficulty ratings of LLM-generated MCQs, the Easy, Medium, and Hard labels correlated closely with its assigned Bloom's Taxonomy levels (Spearman $\rho=0.90$), reflecting an artifact of co-generation. However, neither metric predicted empirical item difficulty (difficulty label $\rho=0.06$; Bloom level $\rho=0.02$). The ratings reflect the structural formatting of a question rather than its underlying difficulty.
Read the full abstract ▾
Fluid Dynamics
A catalog of exact steady states in Kolmogorov flow
Matthew Golden
Two-dimensional Kolmogorov flow is a common proving ground for the dynamical systems picture of
turbulence. Hundreds of unstable periodic orbits have been found for this system across many Reynolds numbers and domain sizes, although the most well explored is Re$=40$ on a square $2\pi\times 2\pi$ domain with a forcing wavenumber of $k=4.$ Here I present a catalog of 8952 exact steady states of the flow in this canonical parameter regime. These solutions provide a fixed constellation informing the dynamics of both
turbulence and transients.
Read the full abstract ▾
Functional Analysis
Adjoints, Order and Spectral Moduli on $L^2(T)$
Mohamed Amine Ben Amor
We study adjoints of $R(T)$-linear, $T$-strongly bounded operators on the conditional space $L^2(T)$ and the interaction between its
Hilbert-type geometry and its Riesz-space order. This produces two natural operator
moduli: the Riesz–Kantorovich order modulus and the spectral modulus $(S^*S)^{1/2}$. After isolating the $T$-regular operators, we show that they form a Dedekind-complete order ideal in the natural
lattice of order-bounded $R(T)$-module homomorphisms. On this operator
lattice the order operations remain $T$-strongly bounded and the adjoint is an order
automorphism. We then compare the two
moduli and explain why they may differ.
Read the full abstract ▾
Group Theory
Hurwitz Equivalence of Reflection Factorizations in Types $B$ and $D$
Patrick Wegener
A
conjecture of Lewis asserts that two reflection factorizations of the same element in a finite Coxeter group $W$ belong to the same Hurwitz orbit if and only if they generate the same subgroup $H$ of $W$ and have the same multiset of $H$-conjugacy classes. We prove this
conjecture for every finite Coxeter group all of whose irreducible components are of type $A$, $B$ or $D$. Furthermore we prove a reduction, which reduces the
conjecture to a finite problem for the exceptional types.
Read the full abstract ▾
Computation and Language
The Illinois Social Attitudes Aggregate Corpus (ISAAC): An Open Tool and Reproducible Pipeline for Analyzing Social Group Discourse at Scale
Babak Hemmatian, Sarah Hadjarab, Jessica Chen et al. (+1 more)
We introduce the Illinois Social Attitudes Aggregate Corpus (ISAAC), an open, modular, and accessible corpus of 527 million+ English-
language Reddit posts selected for relevance to six key social group distinctions based on race, sexuality, age, ability, body weight, and skin tone, covering the 17-year period from 2007 to 2023. A multi-step, human-audited filtering pipeline was used to keep irrelevant content in the curated
dataset below 10%, both overall and for each social group distinction. Each post was then algorithmically annotated with the user's estimated home region, along with a suite of validated off-the-shelf and custom semantic labels including moralization, sentiment,
emotion, and linguistic generalization. We confirm the validity of the resulting corpus through convergent evidence linking ISAAC to macro-level societal trends, such as online search behavior, temporal spikes during major societal events (both nationally and regionally), and long-term shifts in public attitudes. By offering a unified, public infrastructure, ISAAC eliminates research fragmentation and enables seamless replication while supporting diverse empirical workflows at scale. Specifically, ISAAC allows investigators to perform cross-category comparisons, conduct high-precision tracking of long-term temporal shifts in social group discourse, and map spatial variation onto localized public opinion and policy outcomes. ISAAC's fully public, modular pipeline facilitates easy extension of the corpus to new platforms,
languages, and social categories. To accommodate various research needs, ISAAC is accessible both without coding through a point-and-click website and labeler web-apps, and programmatically via an SQL playground, a Python package, and HuggingFace.
Read the full abstract ▾
All authors Babak Hemmatian, Sarah Hadjarab, Jessica Chen, Benedek Kurdi
Software Engineering
Stage-Supervised Latent Reasoning for Single-Shot JavaScript Deobfuscation
Rong Feng and Suman Saha
JavaScript obfuscation is widely used to protect code, but it also makes program analysis and security review substantially harder. Existing LLM-based deobfuscation methods usually treat the task as one-step translation, ignoring the staged structure of practical deobfuscation pipelines. This WIP paper proposes a stage-aware latent
reasoning framework that converts intermediate outputs from a deterministic deobfuscation tool into supervision for Coconut-based training. The model learns from multi-stage rewrites during training but generates the final cleaned program in a single shot at inference time. Preliminary results on JsDeObsBench show that the Coconut-based model improves syntactic validity to 50%, compared with 15% for direct fine-tuning and 25% for a zero-shot baseline, and reaches 80% semantic correctness among valid outputs, indicating better semantic faithfulness than either comparison model in deobfuscation.
Read the full abstract ▾
High Energy Astrophysical Phenomena
Disk truncation triggers relativistic jet launching of a highly accreting supermassive black hole
Hirofumi Noda, Kazuhiro Hada, Hiroto Maruyama et al. (+10 more)
Relativistic jets are among the most energetic phenomena in the Universe and influence
galaxy evolution through energetic feedback, yet the physical connection between
accretion flows and jet launching remains unresolved. Jet production is widely attributed to the Blandford-Znajek mechanism, which extracts black hole rotational energy via large-scale
magnetic fields sustained by an
accretion flow. Although this mechanism is observationally well supported for hot, geometrically-thick
accretion flows, highly accreting environments with strong jets present a paradox, as they are typically dominated by cold, geometrically-thin disks incapable of sustaining large-scale
magnetic fields. Here, we report coordinated X-ray microcalorimeter spectroscopy achieved by X-ray Imaging and Spectroscopy Mission (XRISM) and millimeter/
radio interferometric observations by Global Millimeter VLBI Array (GMVA), Very Long Baseline Array (VLBA), and East Asian VLBI Network (EAVN) on the broad-line
radio galaxy 3C120. High-resolution X-ray spectroscopy reveals a relativistically-broadened Fe-K$\alpha$ line, indicating that the cold, geometrically-thin disk is truncated at $\sim 20~R_{\rm g}$ ($R_{\rm g}$ is the gravitational radius), with the inner region replaced by a hot, geometrically-thick flow. Concurrently,
radio imaging indicates that jet profile extrapolation toward the black hole horizon yields a radius comparable to or even narrower than the disk truncation radius. These results demonstrate that, even at high
accretion rates, jet launching is linked to a hot inner flow, possibly via the Blandford-Znajek process, providing evidence for a universal disk-jet connection via
geometric transition.
Read the full abstract ▾
All authors Hirofumi Noda, Kazuhiro Hada, Hiroto Maruyama, Taishu Kayanoki, Kenji Toma, Taiki Kawamuro, Yoshiyuki Inoue, Tomohisa Kawashima, Soyeon Yeo, Ilje Cho, Keiichi Asada, Yasushi Fukazawa, Yoshihiro Ueda