Artificial Intelligence
Memorizing Who, Knowing How: Benchmarking Personal and Procedural Memory for Persistent Assistants
Maosheng Qin, Qi Zhang, Mingxuan Xia et al. (+5 more)
Memory systems for persistent personal
agents must move beyond isolated
memories, maintaining both personal
memory about whom they serve and procedural
memory about how to act. Crucially, in harness-based personal assistance, procedures or skills are inherently user-conditioned: personal context determines which procedures apply, how they are instantiated, and when they should change. Existing
benchmarks typically emphasize either personal or procedural
memory, leaving their interaction underexplored. We introduce PraxisMemBench, a harness-based
benchmark for persistent personal assistance across daily-life and work trajectories, where personal and procedural
memory evolve together and later tasks depend on both. Experiments show that maintaining two
memory types separately is insufficient, with completion dropping from single-type to interdependent dependencies (84.6% to 77.9%). More broadly, across existing methods, relation-related failures increase from early to late sessions as interaction history grows (59.8% to 72.9%), revealing a continual-maintenance challenge: writing correct
memories is not enough if their evolving content and organization cannot be preserved for later
retrieval. To address this, we propose PraxisMem, which uses privileged joint-state supervision to internalize the ability to maintain
memory content and organization into the
memory manager, with location-aware GRPO and dynamic residual-routed OPSD. PraxisMem reaches 40.78% overall Session Success Rate, compared with 20.60% for Claude Code's native
memory.
Read the full abstract ▾
All authors Maosheng Qin, Qi Zhang, Mingxuan Xia, Jinglei Chen, Jiexiang Wang, Sheng Guo, Cheng Peng, Haobo Wang
Social and Information Networks
Ergodic and Metastable Behavior of Stochastic Popularity Dynamics in Social Networks
Chiara Ravazzi, Paolo Frasca and Gaya Cocca
This paper studies popularity dynamics in social networks, where users' engagement with competing channels is driven by intrinsic content quality, social influence, and popularity feedback. We introduce a
stochastic model in which users interact through a directed influence network and channel exposure depends jointly on these three mechanisms. We first characterize the resulting dynamics in the absence of intrinsic quality, showing that popularity concentrates asymptotically on a single channel, leading to a winner-takes-all regime. When all channels have positive intrinsic quality, we establish ergodicity of the popularity process and characterize its long-run behavior. We further show that, when intrinsic quality is small, the system exhibits metastable behavior: it can remain for long periods near configurations dominated by a single channel before transitioning to another. These results highlight how weak quality differences can coexist with strong path dependence induced by popularity feedback. We then investigate platform-level interventions and propose controls that compensate for network-induced exposure biases and mitigate excessive concentration. Analytical results are illustrated through numerical simulations and an empirical analysis of YouTube data, whose observed popularity dynamics are consistent with the proposed
stochastic framework. The results provide a mathematical framework for understanding the interplay between quality, social influence, and reinforcement in online popularity dynamics.
Read the full abstract ▾
Robotics
MorphAct: Robot Policies Morphing with the Worlds They Act In
Ziyan Liu, Jiting Liu, Weihong Zhong et al. (+7 more)
Progress toward general intelligence in the physical world requires scaling
robot learning across diverse environments and interactions. However, existing
robot policies use shared parameters that remain fixed across interaction contexts. When different operations favor competing parameter adjustments, fitting their mixture can leave individual operations insufficiently adapted. Our atomic-task diagnostics reveal competing update directions in both
vision-
language and action parameters, with source-improving updates increasing loss on other tasks. This motivates MorphAct, a contextual parameterization framework that dynamically adjusts
robot policy parameters as interaction contexts change. Two coupled generators produce context-dependent low-rank updates for the
vision-
language and action modules. Their coupling follows the policy's computational dependency:
vision-
language updates shape the representations that condition action-side parameter generation. Both generators are trained jointly through the native action objective, without operation labels or expert-adapter targets. MorphAct improves performance across three VLA backbones on LIBERO and Meta-World, with gains of up to 16.30%. In real-world evaluation, it achieves gains of up to 14% over $\pi_{0.5}$. Ablations support the benefits of contextual parameter generation and representation coupling.
Read the full abstract ▾
All authors Ziyan Liu, Jiting Liu, Weihong Zhong, Wenda Yu, Yeqiu Chen, Jason Li, Xiaojiang Peng, Shuyuan Hu, Kai Wang, Bo Zhao
Machine Learning
On the Convergence of Semi-Gradient TD: A Mean Field Perspective
Haruki Settai, Tadashi Kozuno and Shinji Ito
Semi-gradient temporal-difference learning is a core algorithmic component of reinforcement learning, but bootstrapping with function approximation generally induces a non-conservative vector field and can lead to instability. We analyze this effect for two-layer neural networks in the mean-field regime and identify a simple stability mechanism in which
entropy-induced relaxation dominates the sensitivity of the moving bootstrap target. Under this condition, we prove exponential convergence to a unique fixed point and quantify its bias relative to the unregularized Bellman solution. We also show that the same mechanism extends to
entropy-regularized off-policy control through soft Q-learning.
Read the full abstract ▾
Machine Learning
ChemArgus: Automated Scientific Diagnostics for Lab-Grade Multimodal Chemistry Reasoning: A Frontier Benchmark
Hanyao Guo, Mo Cui, Zihan Tan et al. (+21 more)
Chemistry evaluation has moved past multiple-choice probes toward open-ended expert problems, yet the
benchmarks that define the current frontier still score answer endpoints alone. Such scoring conflates memorized facts with stepwise derivation, treats visual and textual inputs as interchangeable, and confounds long-horizon planning with stepwise execution. These blind spots prevent any diagnosis of where the
reasoning chain breaks in a model. We present ChemArgus, a competition-level chemistry
benchmark drawn from chemistry competition material, textbooks, and research papers, whose scoring replaces the answer endpoint with three measurement properties. Fine-grained scoring grades every subquestion point by point with partial credit, so a score lands at the level of the step rather than the outcome.
Multimodal capability is priced on every problem containing a real image, paired with a controlled text description, so visual understanding is measured against text on identical tasks. Diagnostic attribution reads paired metrics that share one scoring unit and differ only in context construction, so a gap between two scores isolates a single cause, namely integration, propagation, or local
reasoning. GPT-5.6-Sol scores 23.2% on unassisted whole-problem solving against 28.8% and 30.4% under the two sequential metrics, a gap measuring the integration burden. Through rubric partial credit and these cause-isolating score gaps, ChemArgus shows not only whether a model answers chemistry correctly but where its
reasoning chain fails, so that every score it reports doubles as a diagnosis, a readout of the chain rather than a verdict, aiming to guide
language models step by step on their path toward chemistry expertise, one corrected step at a time, each step priced by the rubric, point by point.
Read the full abstract ▾
All authors Hanyao Guo, Mo Cui, Zihan Tan, Leixin Sun, Yichun Wang, Yinan Shen, Guojun Zhu, Jia Jiang, Erhao Chen, Zui Chen, Wenzhe Gao, Zhihao Yu, Qijun Wang, Yetian Zhu, Jiande Chen, Chuhang Pei, Yao Jiang, Yichen Fan, Pinhan Wang, Xiaorui Pan, Qianyi Zhu, Yang Liu, Mang Ye, Guancheng Wan
Machine Learning
Outperforming 1000 Euclidean Layers With Hyperbolic Goal-Conditioned RL
Timo Klein, Thomas Lang, Andrii Shkabrii et al. (+3 more)
Scaling self-supervised contrastive reinforcement learning (CRL) to 1000-layer Euclidean neural networks has enabled state-of-the-art goal-reaching performance, yet wall-clock time scales linearly with depth. We find that changing the geometry achieves stronger performance with far shallower networks. This matters in complex navigation environments, where walls can make route distance deviate substantially from Euclidean distance. In a U-shaped maze, we show that the longer the route, the more it deviates from Euclidean distance.
Hyperbolic CRL addresses this mismatch by replacing Euclidean distances in CRL's contrastive objective with geodesic distances in
hyperbolic space. On hard navigation tasks, inductive bias beats scale: depth 64
hyperbolic agents outperform 1024-layer-deep Euclidean networks while training more than $12 \times$ faster and using about $13 \times$ fewer parameters. We observe that
hyperbolic agents behave qualitatively better in hard humanoid mazes, where they follow corridors to reach goals, whereas Euclidean
agents try straight-line shortcuts over walls. Our results show that matching the representation geometry to the task can substitute for substantial network depth. Code is at https://github.com/Probabilistic-and-Interactive-ML/HypGCRL .
Read the full abstract ▾
All authors Timo Klein, Thomas Lang, Andrii Shkabrii, Lukas Miklautz, Yllka Velaj, Sebastian Tschiatschek
Machine Learning · Cross-listed from stat.ML
Minimax estimation of the pseudo-spectral gap of continuous Markov chains, with applications to empirical concentration bounds
Vahe Karagulyan, Mikołaj Kasprzak and Pierre Alquier
We study the minimax estimation of the pseudo-spectral gap $\gamma_{ps}$ of a Markov chain from the observation of a single trajectory. Extending earlier works on finite-state space chains by Wolfer and Kontorovich (2024), we provide an
estimator in the case of a compact, continuous state-space. Our
estimator appears to be minimax-optimal on the class of chains with a transition operator with Lipschitz density, up to logarithmic terms. Our
estimator, combined with Paulin's version of Bernstein inequality, allows to derive empirical concentration inequalities for Markov chains. This unlocks many applications, from confidence intervals, to active learning and generalization bounds. In particular, for rested bandits with Markov rewards, we provide a version of UCB with a regret bound similar to the one in the iid case, without assuming any knowledge on the pseudo-spectral gap of the chains.
Read the full abstract ▾
Machine Learning
Provable Transformer Context Compression for In-Context Reinforcement Learning
Yajie Zhao, Yingjun Mo, Zhiyong Wang et al. (+3 more)
Transformers can perform in-context reinforcement learning without parameter updates. However, existing theoretical constructions often require context sizes that grow with the number of sampled transitions, leading to increasing
memory costs in large-scale problems. This raises a fundamental scalability question: how much context is intrinsically necessary to carry out such learning? We answer this question for fixed-policy TD evaluation. We show that, with a shared linear decoder, the minimum context dimension required to exactly reproduce the first $L$ TD iterates is the Krylov dimension induced by the sampled transitions. When the available context dimension $m$ falls below this threshold, exact reproduction is no longer possible. In this regime, we characterize the optimal approximation error and show that it decays as $\Theta(\gamma^m\max\{1,1/(m(1-\gamma))\})$ where $\gamma$ is the discounting factor. Moving from representation complexity to actual
memory cost, we characterize the optimal retained
memory for exact continuous representations as $\Theta(nk_L)$ and show experimentally that on a sampled task whose exact rank is full but whose numerical rank is small, a context accurate to $10^{-2}$ is $88.7\%$ smaller than the sparse sufficient statistics of TD. Finally, we establish the learnability of the shared executor through pretraining by proving global convergence of an auxiliary objective over effective
attention parameters under suitable task diversity. Experiments across diverse TD evaluation problems validate parameter recovery and the predicted context–accuracy tradeoff.
Read the full abstract ▾
All authors Yajie Zhao, Yingjun Mo, Zhiyong Wang, Bo Tang, Chao Xu, Fang Kong
Artificial Intelligence
What Coding Agents Leave Unsaid: Measuring Task-State Gaps Between Execution and Reports
Guoyang Xu, Rui Zhang, Xiuyu Wu et al. (+1 more)
Human supervisors rely on coding
agents' reports to assess progress, verify repairs, and decide when to intervene. But a supervisor cannot investigate a failure they were never told occurred. Failed attempts, workarounds, and unresolved risks may disappear from routine reports, leaving supervisors unaware of issues that warrant scrutiny. We conduct CTXGAP, a measurement study quantifying the gap in task-state recoverability between
agents' accumulated user-facing reports and their full recorded execution contexts. Across 37 coding tasks and 4,799 valid probes,
agents answer 93.6% of task-state questions on demand; full-context readers recover 92.7%, while report-only readers recover 52.0%, a 40.7-point gap. At the final checkpoints of 30 verified fixes, the gap remains 34.8 points. Human annotation identifies states mentioned without enough supporting detail as the largest category of unrecovered information, and model-assessed supervision-relevant states retain substantial gaps. Full terminal scrollback improves recovery to 73.6% but still trails full context. Under matched report lengths, context-informed rewrites improve recovery by about eight points, without a detectable targeted-over-natural advantage. Positive gaps persist across harnesses and independent DeepSeek execution and measurement. CTXGAP establishes report-based task-state recoverability as a complementary evaluation dimension.
Read the full abstract ▾
All authors Guoyang Xu, Rui Zhang, Xiuyu Wu, Hao Chen
Computation and Language
A Strong Harness Narrows the Model Gap: A Comparative Evaluation for Legislative Document-to-XML Conversion
Palvika Bansal, Rittika Jindal and Ruichong Wang
Legal-content reproduction requires matching the source exactly in section hierarchy, verbatim text, citations, and amendment spans; a dropped citation or mislabeled amendment carries into the products lawyers rely on. We report an industrial pipeline converting U.S. state legislation into strict, schema-constrained XML, and hypothesize that its accuracy comes from harness engineering, the deterministic tooling built around a model, more than the model itself. Our controlled ablation crosses three harnesses (zero-shot, a single tool-using
agent, and a decomposed multi-
agent pipeline) with three Claude models on 153 gold bills from two jurisdictions. The harness narrows the model gap: the weakest model in the full pipeline beats the strongest model alone (0.948 versus 0.806, paired bootstrap p less than 0.001). Because the two agentic harnesses share identical tools, their gap comes from orchestration: decomposition plus the persisted state and per-
agent context it brings. Across four scored dimensions, the model spread at zero-shot (0.68 to 0.81) narrows to 0.95 at the full pipeline; the residual sits almost entirely in citation normalization, where the canonical identifier is retrieved from an external authority rather than generated from the weights, so a deterministic lookup tool supplies the exactness that a larger model does not.
Read the full abstract ▾
Human-Computer Interaction
ShadAR: Real-time Prompting On-demand Visual Augmentation in Augmented Reality
Yanni Mei, Samuel Hartenstein, Jonas Wombacher et al. (+2 more)
Augmented Reality can reshape our visual perception (e.g., simulating color blindness) yet such effects are typically pre-programmed by experts, making visual augmentation static and inflexible. We introduce ShadAR, an AR system that allows users to describe expected visual augmentations (e.g., "Blur all people") for changing their
vision in real time. ShadAR can graphically manipulate AR passthrough in real time via shader injection. When receiving a user's
prompt, ShadAR uses LLMs (GPT5.2) to interpret the user's intent, analyzes the user's view with object detection (YOLOv26), generates a pixel shader, and applies it to the targeted region on AR passthrough. To understand usage patterns, we invited 21 HCI researchers to use ShadAR, identifying four application domains (design, health, productivity, and social interaction) and two
prompt-framing patterns: procedural
prompts specifying what to modify and how ("Highlight the book."), and end-goal
prompts conveying desired mental or behavioral states ("Make me laugh."). Such end-goal
prompts suggest a future in which users specify outcomes (e.g., "Make me not miss my flight") while the system determines how to realize them, though often with inherently indeterminate results. We discuss design opportunities for ShadAR and ethical concerns.
Read the full abstract ▾
All authors Yanni Mei, Samuel Hartenstein, Jonas Wombacher, Florian Mueller, Jan Gugenheimer
Machine Learning
Empirical Boundaries of Reservoir Computing for Few-Shot Sequence Pattern Tasks
Mario Mallea, Angela Nebot and Franscisco Mugica
Efficient sequence modeling remains a fundamental challenge in machine learning. Despite the success of
Transformers, recurrent neural networks (RNNs) have recently re-emerged as efficient alternatives. Particularly, Echo State Networks (ESNs) offer a compelling framework of randomized RNNs that have been extensively studied for time-series forecasting and dynamical systems. This paper presents an empirical analysis of distributional, architectural, and optimization variants of ESNs evaluated on few-shot sequence-to-sequence (seq2seq) pattern tasks. These synthetic tasks utilize discrete symbol sequences, characterized by delays, inactive phases, and triggered markers to isolate and evaluate
memory and pattern recognition capabilities. The minimality of these tasks is crucial to understand the strengths and limitations of machine learning models. Furthermore, when the input signal utilizes a one-hot encoding, ESN dynamics are governed strictly by their weight distributions, offering an ideal lens through which to study reservoir computing. Our findings map the performance boundaries of ESNs relative to state-of-the-art (SOTA) models. We show that preserving a linear state regime is vital for
memory tasks, such as pattern completion or simple copy. We propose a data-dependent modulation mechanism for nonlinear input injection into the reservoir (G-ESN). G-ESN functions as a statistical filter capable of solving tasks that conventional ESNs cannot (the adding problem and bracket matching). Finally, for Deep ESNs, we evaluate a shared readout mechanism across layers that outperforms conventional concatenated readouts on structural tasks, including selective copy, sorting problem, and bracket matching.
Read the full abstract ▾
Computation and Language
Gray-Box Hallucination Detection in LLMs with Token Identities
Shaked Goldshtein, Yitzchak Grunbaum, Fabrizio Frasca et al. (+2 more)
Large
language models often produce fluent but factually incorrect output. Gray-box detectors identify such
hallucinations from the
token-probability statistics emitted during generation, avoiding the need for white-box access to hidden states that frontier models do not provide. These detectors, however, typically rely only on probability-derived representations and ignore the identities of the candidate
tokens. They can therefore observe how uncertain the model is, but not what it is uncertain about. Consequently, they cannot distinguish between a model that is torn between "big" and "small" and one that is torn between "big" and "huge" when the two pairs receive the same probability distribution. We introduce GURI (Gray-box Uncertainty Representations with Identity), a lightweight gray-box
hallucination detector that combines candidate-
token probabilities with per-
token vector identifiers. This exposes which alternatives the model considered while preserving the accessibility and transferability of the gray-box setting. Across three open-weight models and three question-answering
benchmarks, GURI increases mean test ROC-AUC from 72.09% to 79.09% over a probability-only baseline. On the closed-source GPT-4.1 model, mean ROC-AUC increases from 67.39% to 71.84%. Among the three open-weight models, GURI outperforms transferred probability-only baseline in 17 of 18 directed cross-model transfers without fine-tuning. For the three open-weight models, cross-
dataset transfer between HotpotQA and TriviaQA with the model held fixed improves mean ROC-AUC over transferred LOS-Net by 3.76 points.
Read the full abstract ▾
All authors Shaked Goldshtein, Yitzchak Grunbaum, Fabrizio Frasca, Guy Bar-Shalom, Haggai Maron
Computer Vision
Why 3D Reconstruction Breaks Down on Under-Ice AUV Video, and How to Predict It
Ana-Ştefania Kogălniceanu, Klára Orbán, Ashwin Nedungadi et al. (+2 more)
Reconstruction of 3D structures using under-ice AUV
video footage has been inaccurate, although it is not certain whether this is due to shortcomings in the methods or an inherent deficiency of the capture system. We test four different models for reconstruction on the same Antarctic under-ice AUV
video (Icefin, Kamb Ice Stream): classical feature matching SfM, differentiable photometric splatting, volumetric depth fusion, and a pose-free feed-forward foundation model. All four fail, each in a distinct way: near-total registration yielding almost no structure, geometrically broken output despite healthy photometric metrics, a persistent coverage artifact, and unstable cross-chunk scale. We show that all four are consistent with the same capture regime, a small camera-baseline-to-scene-depth ratio. Following our analysis, we propose a detection system that flags this type of failure in minutes, rather than hours of wasted compute.
Read the full abstract ▾
All authors Ana-Ştefania Kogălniceanu, Klára Orbán, Ashwin Nedungadi, Stefan Lüdtke, Kuderna-Iulian Benţa
Computation and Language
SoK: Grading Open-Text Legal Answers, from Law School Exams to Trustworthy Rubric Judges
Sebastian Nagl
Law students are examined through open-text answers: case solutions, essays and memos that apply the law to a set of facts. Such answers are expensive to grade, human graders disagree widely on them, and large
language models (LLMs) now offer to grade them at scale, both in teaching and in the
benchmarks that test legal AI. This paper systematizes how open-text legal answers can be evaluated. We derive eight desiderata from educational measurement and from the law that governs exams, among them validity, reliability, reasons, credit for defensible alternative solutions, robustness to gaming and human oversight. We then compare eight evaluation paradigms against them, from unstructured human grading to LLM judges with instance-specific rubrics and pairwise comparison. Three findings stand out. A rubric written for the task is the strongest lever for agreement between an LLM judge and human graders, but the reported figures answer different questions and rarely come with the human agreement they must be read against. The central validation gap lies where examination law is most demanding: defensible alternative solutions, follow-on errors, and the examiner's correction of model judgments. And rubric-centred grading is the most auditable family, because it makes the standard, its application and the final decision checkable. From this we derive the trustworthy rubric judge: an instance rubric that credits arguments and alternative paths,evidence checked in code, probes, repeated runs, calibration against independent human grades,logging, and an examiner who reads every script. We state what each safeguard rests on, what remains to be tested, and the uses the evidence supports today.
Read the full abstract ▾
Human-Computer Interaction
Think Once, Reflect Twice: How Feedback Shapes Human-AI Collaboration
Joshua Holstein, Philipp Spitzer, Sebastian Schäfer et al. (+2 more)
As artificial intelligence (AI) increasingly supports human decision-making in high-stakes domains, designing effective human-AI collaboration becomes critical. Successful collaboration requires humans to reflect on decision-making processes and adjust their strategies accordingly. Current research typically investigates interventions, such as explainable AI, that target reflection during the decision-making process (reflection-in-action). Less
attention has been given to retrospective examination of decisions (reflection-on-action). This work explores reflection-on-action by studying the impact of feedback on reliance behavior and human-AI team performance. Specifically, we investigate feedback interventions targeting overrides of AI advice through a behavioral study (N=244). We find that such feedback influences reflection-on-action, while unidirectional feedback risks reducing subsequent reflection-in-action. Furthermore, we identify opposing shifts in reliance behavior that leave overall human-AI team performance unchanged but show context-dependent effects. Balanced feedback did not show these risks, preserving both concurrent and retrospective reflection, offering insights for designing for appropriate reliance on AI.
Read the full abstract ▾
All authors Joshua Holstein, Philipp Spitzer, Sebastian Schäfer, Maximilian Förster, Gerhard Satzger
Computational Engineering
Transverse impulse response of finite and infinite Rayleigh beams
Saranya Biswas and Anindya Chatterjee
Here, we study an undamped Rayleigh beam subjected to a concentrated transverse impulse applied at a point far from the boundaries. For a simply supported finite beam struck at its midpoint, using the modal expansion of the impulse response, a short-time asymptotic series for the response of the struck point is derived. The infinite beam impulse response is obtained as a spectral integral in two ways. The spectral integral is evaluated exactly in terms of modified Bessel functions via Laplace transform calculations. The corresponding short- and long-time asymptotics of the exact integral are consistent with the power series developed earlier and the Euler-Bernoulli beam, respectively. Direct numerical evaluation of the integral with a pre-specified error bound and numerical contour integration is also presented for verification. An ad hoc approximation is given which matches the response well for all times. The relationship between finite and infinite beam responses is examined in some detail. There are three consequences of finite length. First, the coefficients in the power series expansion contain components that decay exponentially with increasing length of the beam. Second, a smooth deviation from the infinite beam response is observed, consistent with the greatest group velocity in the beam, wherein a small precursor is detected which is consistent with broader theory. Third, a slope discontinuity is observed, corresponding to the high-frequency limit of group velocity. The full time series plot from modal expansion satisfactorily shows both the arrival and passage of these high frequency components. This paper opens the door to impact studies that are envisaged in future work.
Read the full abstract ▾
Numerical Analysis
Spherical Perfectly Matched Layer Method for Electromagnetic Scattering Problems with Large Wave Number
Jian Shi, Yonglin Li and Haijun Wu
We study a spherical perfectly matched layer (PML) method for the time-harmonic Maxwell scattering problem with large wavenumber. Using vector spherical harmonics and a modal transformation for the coupled transverse
magnetic components, we establish the sharp wavenumber-explicit stability estimate for the truncated PML problem and prove that its inf-sup constant is of order $O(\kappa^{-1})$. We then analyze the first-order Nédélec edge element method and its continuous interior penalty counterpart. Under the condition that $\kappa^3h^2$ is sufficiently small, the preasymptotic errors in the energy norm and the $\kappa$-scaled $\mathbf{L}^2$ norm are of orders $O(\kappa h+\kappa^3h^2)$ and $O(\kappa^2h^2+\kappa^3h^2)$, respectively. Numerical experiments confirm the theoretical results.
Read the full abstract ▾
Multimedia
Shared Sentiment Evidence from Frozen Multimodal Models for Continuous Sentiment Prediction
Shuobo Wang, Ruiqi Chen, Shaobo Qiao et al. (+2 more)
Continuous
multimodal sentiment analysis estimates polarity and intensity from linguistic, acoustic, and visual cues that may disagree. Hard judgments compress sentiment alternatives, while learned fusion interfaces tie task adaptation to model representations. We introduce Shared Sentiment Evidence (SSE), an output-space interface with fixed view and sentiment-option semantics across frozen backbones. Four queries yield five-option distributions whose probabilities,
entropy, and cross-view relations form a 28-dimensional vector for CatBoost regression. On CH-SIMS-v2 and CMU-MOSI, the Qwen-based interface improves on supervised joint-only readout. Matched Ridge and CatBoost controls assess the contributions of additional views, deterministic summaries, and nonlinear readout. With the same interface specification and fixed readout settings, separately fitted heads also improve on joint-only prediction for three additional frozen backbones. These results support learning continuous sentiment from reusable output coordinates.
Read the full abstract ▾
All authors Shuobo Wang, Ruiqi Chen, Shaobo Qiao, Shixu Yang, Feng Xu
Computational Geometry
Towards Learning-Guided Discovery of Extremal Geometric Constructions: A PatternBoost Study
Deepak Ajwani, Paula Carroll, Rishikesh Gajjala et al. (+5 more)
Extremal constructions play a central role in the design and analysis of
algorithms. Integrality-gap instances expose the limitations of relaxations, while hard configurations reveal structural sources of computational difficulty. Such constructions, however, are still typically discovered through a slow, manual process. This paper investigates how machine learning can accelerate their discovery to support
algorithm design and analysis. We build on PatternBoost, a framework alternating local search with generation from a trained
transformer, using its Axplorer implementation to investigate three problems: integrality gaps for maximum independent set of rectangles, integrality gaps for stabbing unit squares with axis-parallel lines, and hard instances for recursive guillotine cutting. By exploring candidate representations, local search operations, and surrogate models used to rank candidates before exact verification, we carefully adapt the framework to these combinatorial geometry problems.While naive representations and search strategies fail to find non-trivial instances, our tailored adaptation discovers numerous non-trivial constructions. For stabbing unit squares, the method reaches a gap of $20/13\approx 1.54$, improving on the literature's best known lower bound of $3/2$. Purely automated search converges rapidly and then plateaus, prompting the question of what is required to push the bound higher still. We find that initial instance diversity plays an important role: with modest expert input via a structured staircase construction, the system reaches an improved integrality gap of $14/9\approx 1.555$. These results demonstrate that machine learning and domain expertise can be combined to discover new extremal constructions that advance algorithmic lower bounds.
Read the full abstract ▾
All authors Deepak Ajwani, Paula Carroll, Rishikesh Gajjala, Ananta Manoranjan, Ryan O'Connor, Rajiv Raman, Saurabh Ray, Aimen Taha
Human-Computer Interaction
AdaptaFace: Adversarial Assistance for Inclusive Facial Gesture Interaction
Siyu Zhang, Yelu Gu, Xiaoyun Xu et al. (+2 more)
Facial gestures can enable access for people with motor impairments (PwM), yet recognizers may not reliably interpret individual facial movement patterns. Existing approaches either constrain agency by requiring predefined templates or impose repeated calibration, data collection, and computation through user-specific retraining. We present AdaptaFace, an input-level compatibility layer using bounded, proxy-dependent
adversarial perturbations to translate user-compatible gestures into inputs for an unchanged facial action unit (AU)-based recognizer. Designed around six personalization principles informed by interviews with 12 PwM, AdaptaFace supports both recognition enhancement and gesture vocabulary extension. In a within-subject study with 16 participants, AdaptaFace increased task completion from 59.7% to 90.5% and reduced workload from 5.61 to 4.21. Offline validation showed strong steering with aligned perturbation generation and evaluation but weaker cross-recognizer transfer, defining the approach's boundary. These findings frame
adversarial assistance as input-side accessibility mediation for personalized facial-gesture interaction.
Read the full abstract ▾
All authors Siyu Zhang, Yelu Gu, Xiaoyun Xu, Kirsten Cater, Oussama Metatla
Systems and Control · Cross-listed from math.OC
L1 Minimum Energy Estimation of Positive Systems
Alba Gurpegui Ramón
We present a novel class of minimum energy estimation problems for positive linear systems with linear costs subject to coupled process and measurement disturbances, where the latter is bounded by the state and the former is bounded by the noise. Through dynamic programming we show that the problem admits explicit solutions. Remarkably, the Luenberger observer structure of the optimal
estimator emerges from the optimization itself, with a switching gain that inherits the sparsity of the disturbance constraints, and it is obtained from a differential equation in the finite horizon case. In the infinite horizon case, an algebraic equation characterizes the limit of the finite-horizon solution and yields a stationary
estimator with a static gain. We introduce two distinct notions of detectability to characterize when the static gain stabilizes the error dynamics while preserving their positivity, and analyze under which conditions the stationary solution can be formulated as a linear program. We further establish that a disturbance perturbation of finite L1-norm yields an estimation error of finite L1-norm. The framework is illustrated on a large-scale irrigation network.
Read the full abstract ▾
Machine Learning
LatentLoop: Scaling Looped MoE Models with Asymmetric Compute Allocation
Zefeng He, Yunkai Chai, Xiaoyang Chen et al. (+5 more)
Looped Mixture-of-Experts (MoE) models combine parameter reuse with sparse computation to improve performance under matched budgets. Prior work demonstrates that different loop visits can play different roles even when they share parameters. To probe the contributions of different loop visits, we learn separate residual coefficients in a model that applies a shared block twice, observing a smaller scale in the first visit. Fixed coefficient experiments show that reducing the first visit's scale preserves or improves performance, whereas the same reductions to the second hurt it. This asymmetry motivates us to explore asymmetric compute allocation. We propose LatentLoop, which compresses the first visit's sequence, retains full resolution in the second, and reallocates saved computation to increase the hidden size. We further learn how strongly the first visit updates the full sequence, and share MoE weights across visits while using separate
attention parameters to accommodate their different sequence resolutions. Across 15 configurations spanning 1.3B–7B parameters, with the largest budget exceeding $10^{21}$ FLOPs, LatentLoop consistently lowers validation NLL. At a fitted NLL of 1.988, predicted compute savings are 19.0% and 7.9% relative to vanilla MoE and the looped baseline. It also improves average scores on nine
benchmarks at all sizes in 0-shot and 5-shot settings. These results suggest asymmetric compute allocation is a promising direction for looped MoE models.
Read the full abstract ▾
All authors Zefeng He, Yunkai Chai, Xiaoyang Chen, Tong Zhu, Zongliang Wu, Yun Luo, Yafu Li, Yu Cheng
Hardware Architecture
STISA: A 0.16 GOPS/W/PE Single-Shot Inference FPGA-based SNN Accelerator with Algorithm and Hardware Co-Design
Kainan Wang, Chengyi Yang, Lei Liu et al. (+4 more)
Spiking neural networks (SNNs) offer great potential for low-power intelligent computing owing to their event-driven nature, yet their deployment is limited by high inference latency and inefficient hardware utilization caused by temporal dependencies. Existing accelerators struggle to balance timesteps, accuracy, and energy efficiency, while hardware-oriented timestep compression remains underexplored. To address these challenges, this work presents STISA, a unified
algorithm-hardware co-design framework featuring: (i) a Temporal Splitter Compression (TSC) technique with a Temporal Split Structure (TSS) to reduce temporal redundancy while preserving network dynamics; (ii) a hardware-efficient spatiotemporal parallelism scheme, which co-optimizes dataflow and architecture through a Temporal-Prioritized Output-Stationary (TPOS) mapping and a flexible Block-Mux Streaming (BMS) architecture; and (iii) a joint TSC-TPOS-BMS co-optimization framework featuring a bandwidth-aware resource allocation strategy to achieve balanced accuracy, latency, and energy efficiency across diverse network architectures. On the algorithmic side, TSC reduces synaptic operations by 24.44%-51.71% while achieving competitive accuracies of 96.38%, 81.10%, 68.51%, and 82.40% on CIFAR-10, CIFAR-100, ImageNet, and DVS-CIFAR10, respectively. On the hardware side, FPGA prototypes deliver up to 7.99 speedup, 30.4% power reduction, and 30.1% LUT reduction. Compared with state-of-the-art SNN accelerators, STISA achieves 0.16 GOPS/W/PE, demonstrating its scalability and superior energy efficiency for real-time SNN inference under constrained hardware resources.
Read the full abstract ▾
All authors Kainan Wang, Chengyi Yang, Lei Liu, Chengting Yu, Yee Sin Ang, Bo Wang, Aili Wang
Cryptography and Security
Diagnose Before You Refine: Multi-Dimensional Judge Feedback for Query-Efficient Jailbreak Attacks
Yixin Tan, Chia-Yi Hsu, Jiayang Liu et al. (+2 more)
Automated jailbreak attacks rely on evaluation feedback to optimize
adversarial prompts, but existing methods often lack fine-grained diagnostic information. We propose Jev-Guided Strategy Bandit (JSB), a query-based jailbreak framework that leverages multi-dimensional probabilistic feedback to guide
prompt optimization. JSB combines continuous jailbreak scoring with targeted refinement, strategy switching, and best-state reseeding to balance exploration and refinement. We evaluate JSB on two harmful request
datasets across five open-source and three closed-source LLMs. Under a budget of only 10 target-model queries, JSB achieves an average HarmBench attack success rate of 89.8% across ten open-source settings, consistently outperforming existing automated jailbreak attacks. Further analyses demonstrate its effectiveness under varying query budgets and attacker models. Defense experiments also reveal limitations in existing filtering mechanisms and highlight the potential of Jev-based output filtering.
Read the full abstract ▾
All authors Yixin Tan, Chia-Yi Hsu, Jiayang Liu, Chia-Mu Yu, Rui Wen
Software Engineering
Temporal Adaptive Fine-Tuning for Code Comprehension
Alex Wolf, Pooja Rani and Harald C. Gall
Repository-level code comprehension question answering (QA) supports developer
reasoning over functionality and usage distributed across codebases. However, repository-specialized large
language models (LLMs) can become misaligned as code evolves. We study this challenge as temporal repository adaptation: acquiring new knowledge, suppressing obsolete, retaining still-valid, and preserving semantically neighboring knowledge. Existing
retrieval, fine-tuning, and model-editing methods address only parts of this problem. We propose Temporal Adaptive Fine-Tuning (TAFT), a novel adaptation approach that treats repository transitions as supervision. TAFT derives temporally grounded QA data from repository transitions, pairs newly valid with obsolete knowledge through preference optimization, and replays transition-conditioned retention examples to preserve still-valid knowledge. We evaluate TAFT on an augmented temporal SWE-bench
dataset comprising 255,295 QA entries across ten temporal slices, comparing deferred (Baseline), per-slice retrained (Baseline-Batched), and variants with knowledge anchors along four dimensions: updated knowledge, obsolete-answer suppression, retention, and neighbor stability. TAFT consistently outperforms both the deferred and batched baselines, with an average relative improvement of 18% in updated/suppressed knowledge accuracy, doubled retention accuracy, and roughly 1.5 times the neighbor accuracy. With directly supplied knowledge anchors, TAFT-KANT further improves mean updated/suppressed, retention, and neighbor accuracy over TAFT by 3.55%, 8.35%, and 13.00%, respectively. Overall, TAFT demonstrates its effectiveness across all evaluated areas, effectively enabling code comprehension QA
alignment with repository
evolution.
Read the full abstract ▾
Networking · Cross-listed from eess.SP
Toward Motion Capture Textiles: A Batteryless IMU Array Powered and Networked Through a Conductive-Fabric Bus
Ryusei Miyakawa and Akihito Noda
Motion capture is increasingly important in healthcare, rehabilitation, and sports applications. However, conventional
vision-based systems require camera infrastructure and line-of-sight conditions, limiting their use in daily environments. This paper proposes a location-independent platform for motion capture, based on an array of IMU (inertial measurement unit) sensor nodes without individual wiring or onboard batteries, enabled by two-dimensional communication and power delivery over conductive fabric. Pipeline processing in each sensor node, combined with a
token-passing protocol, enables collision-free and stable data acquisition at 50 Hz. Real-time three-dimensional reconstruction of flexible objects is further demonstrated by applying Hermite spline interpolation to orientation data from grid-aligned IMUs. Against a cylinder of known radius, the reconstructed surface deviates by 2.60 mm RMS over the 225 mm span of the array. The reconstruction is formulated for the quasi-static regime, and this letter empirically validates it in that regime. The hardware platform and communication protocol implemented in this work enable further investigation toward dynamic motion capture textiles.
Read the full abstract ▾
Machine Learning
Safe-FCL: Continual Risk Awareness and Fine-Grained Safety Aggregation for Federated Continual Learning of LLMs
Shuaijie Ma, Yongkang Hu, Yu Cheng et al. (+4 more)
Federated fine-tuning enables large
language models (LLMs) to learn collaboratively from distributed client data, but malicious client updates can compromise global model safety. As
federated tasks evolve, changing malicious update patterns challenge the ability of existing defenses to maintain reliable risk detection over time. To address this challenge, we propose Safe-FCL, a framework integrating continual risk awareness with fine-grained safety aggregation for
federated continual learning of LLMs. Specifically, the Continual Risk Awareness (CRA) mechanism learns emerging risk patterns across sequential tasks and constrains detector updates using a protective subspace derived from historical features, mitigating forgetting of previously learned malicious patterns while adapting to new risks. To translate risk awareness into aggregation-level safety control, the Fine-Grained Safety Aggregation (FGSA) mechanism combines client-level risk assessments from CRA with module-level deviations from historical update behavior to adaptively regulate aggregation contributions, suppressing potentially malicious updates while minimizing unnecessary restrictions on benign updates. Extensive experiments demonstrate that Safe-FCL outperforms current SOTA methods, improving global-model Rule and MD-Judge scores in
federated continual learning by 0.41% and 0.64%, respectively, while increasing average risk recognition accuracy by 2.01%.
Read the full abstract ▾
All authors Shuaijie Ma, Yongkang Hu, Yu Cheng, Jiuan Zhou, Yi Xie, Yushuo Zhang, Zhaoxia Yin
Computation and Language
Choosing the Number of Attention Heads from Intrinsic Dimension: A Small-Scale Study
Arjun Boya, Charan Sai Bandaru, Masthanaiah Pillimetla et al. (+1 more)
The number of
attention heads in a
transformer is usually set by convention or grid search. We test a simple heuristic for setting it in advance: estimate the intrinsic dimension of
token representations without labels, and use roughly that many heads. We estimate intrinsic dimension with a nested-dropout VAE trained on DistilBERT
embeddings of WikiText-2, choosing the KL weight at the knee of the VAE's rate-distortion curve rather than by hand. This gives $k^*=9.2$ (95% CI [8.6, 9.8]). In a small 4-layer
transformer trained from scratch, 10 heads and 32 heads are statistically equivalent on SST-2 and on MRPC (paired TOST over 10 seeds, $\pm1$ point margin), with 69% fewer
attention parameters for the 10-head model. Training 10 heads from the start also slightly outperforms training 32 heads and
pruning to 10 (0.6 points, $p=0.008$). The evidence is narrow: two classification tasks, one small architecture, no pretrained models, and no formal link between intrinsic dimension and head count. We present the result as a practical heuristic worth testing at larger scale, not as a general principle.
Read the full abstract ▾
All authors Arjun Boya, Charan Sai Bandaru, Masthanaiah Pillimetla, Kavana Chidanandappa
Software Engineering
Taming the Online Software Engineering Tongue: Detecting and Reducing Offensive Language in Software Communities
Jithin Cheriyan, Bastin Tony Roy Savarimuthu and Stephen Cranefield
Online Software Engineering (SE) communities such as Stack Overflow have become unwelcoming and hostile spaces, as evidenced by the presence of rude and offensive comments that are detrimental to the emotional wellbeing of participants. The main objective of this work is to investigate the prevalence and nature of offensive
language in online SE communities, and develop approaches to classify them with the ultimate goal of reducing offensive
language by proposing an Offensive
Language Reduction System (OLRSSE) that has the ability to detect, classify, and explain to the poster why a comment is offensive and also propose non-offensive alternative suggestions. Using data from 130,000 user comments on discussions held in four prominent SE platforms: Stack Overflow, GitHub, Gitter and Slack, we developed six sets of twelve (72 in total) different machine learning models (both traditional and deep learning). The F1-Score of the best performing models in the system ranges from 0.56 to 0.94, with an average F1-Score of 0.85, predominantly using BERT as the primary model for various classification tasks. Also, two
language-based models were developed for paraphrasing offensive to non-offensive comments. The overall utility of the OLRS-SE was evaluated using user studies with 30 participants, which showed that the developed system is easy to use and also effective in reducing offensive
language. Investigating the capabilities of different large
language models for offensive
language reduction is a promising avenue for future research.
Read the full abstract ▾