Distributed Computing
Distributed Service Orchestration in Edge-Cloud Continuum for Digital Healthcare
Johirul Islam, Hafiz Faheem Shahid, Ijaz Ahmad et al. (+3 more)
Today's digital healthcare services rely on various applications and functions that must be continuously accessible. Cloud computing enables global access to these services through public networks, which often also introduce increased latency, higher bandwidth consumption, and additional security risks compared to local operation. Edge computing mitigates these challenges by deploying cloud services closer to the end users and data sources, thereby improving resilience to network disruptions and reducing latency, bandwidth usage, and exposure to security threats. However, service deployment typically relies on the availability of centralized registry servers. Consequently, if the network connection or the registry server itself becomes unavailable, service deployment at the target edge node may fail. To address this, we propose a three-tier registry architecture to enhance deployment reliability and service availability, considering DockerHub as a remote public registry, an MEC-based off-premises registry as a remote private registry, and a LAN-based on-premises registry as a local private registry. The performance and efficiency are analyzed through measurements related to the estimated deployment time, inflicted network and computational load, and energy consumption, while the required nanoservices are deployed from different tiers. The experimental results indicate that alongside the improved tolerance to network disruptions, the local private and remote private registries also outperform the remote public registry in the deployment performance. The findings highlight the effectiveness of proximity-aware service distribution in improving the resilience, performance and efficiency of service deployment.
Read the full abstract ▾
All authors Johirul Islam, Hafiz Faheem Shahid, Ijaz Ahmad, Tanesh Kumar, Ayan Mondal, Erkki Harjula
Computation and Language
Persuaded, Not Informed: Incentive-Misaligned Witnesses Defeat In-Context Grounding
Rahul Balakavi
Language-model
agents increasingly answer questions over customer-relationship management (CRM) records, such as whether to qualify a sales lead. We identify a failure mode not addressed by a stronger model: when the context contains an assertion by a party with an incentive toward optimism - here the sales representative, a witness recorded in the CRM - the model treats the assertion as evidence and clears deals the company's own records deem unacceptable. Across 100 lead-qualification tasks from CRMArena-Pro, the representative asserts an acceptable timeline in every call and an acceptable budget in 76; on the 31 tasks where such an assertion contradicts the price list and installation policy, a model reading only the transcript clears the deal in 29 of 31 cases. The signature is consistent across seven models from four providers (misled on 87-97%); scale and explicit
reasoning confer no resistance. Only 3 of 35 genuine failures involve no assertion: the failure is persuasion, not missing information. We contribute a diagnostic method rather than an architecture: (i) a bucket analysis that separates persuasion from information gaps, (ii) a same-information control showing that supplying the records to the model lowers strict accuracy from 41 to 18 while raising recall - precision collapses - and (iii) a compute-step control that holds extraction fixed and varies only who computes Budget and Timeline. The margin ranges from 42 points on an inexpensive model to 2-5 points on models that already compute correctly; on the strongest models the arms are within confidence intervals, so the pattern is a consistent direction and a soundness property, not a proved performance floor. We pre-specify a generalization test that returns a negative result, characterize the precondition (a policy exactly specified in the inputs), and release all evaluation artifacts.
Read the full abstract ▾
Networking
Has The Physical Layer Matured?
Mansoor Shafi, Changlong Xu, Xingqin Lin et al. (+5 more)
The wireless physical (PHY) layer has enabled successive generations of cellular systems through advances in modulation, coding, waveforms, and multiple-input multiple-output (
MIMO) transmission. This article assesses whether these techniques are now approaching maturity and where substantial further gains remain possible. Field measurements and quantitative evaluations indicate that many link-level refinements, including constellation shaping, channel-code
evolution, reduced-complexity receivers, and waveform enhancements, remain valuable but typically provide bounded gains that must be balanced against implementation complexity and overhead. In contrast, massive
MIMO and distributed
MIMO offer a more scalable system-level opportunity by increasing the number and quality of usable spatial channels. Their effectiveness relies on time-division duplex reciprocity for scalable channel state information (CSI) acquisition, while practical limitations include calibration, pilot reuse, channel aging, weak pilot reception from
cell-edge users, and the fronthaul and synchronization requirements of coherent distributed operation. Artificial intelligence and machine learning (AI/ML) provide complementary opportunities for further PHY layer innovations. The PHY layer is therefore not dead; its most consequential advances will come from deployable spatial processing and CSI acquisition, complemented by targeted link-level and AI/ML-based refinements.
Read the full abstract ▾
All authors Mansoor Shafi, Changlong Xu, Xingqin Lin, Gilwon Lee, Feifei Sun, Eko Onggosanusi, Oskari Tervo, Joonyoung Cho
Computer Vision
Looks the Same, Answers Differently: Flip-Direction Steering for Robust Vision-Language Reasoning
Yeonsung Jung, Joonhyun Jeong, Hoang Pham et al. (+4 more)
Vision-
language models (VLMs) achieve strong visual
reasoning performance, yet subtle changes from routine image capture and processing can alter their
reasoning trajectories even when images appear nearly identical. In long-horizon generation, the resulting activation shifts may accumulate across decoding steps, progressively altering
reasoning tokens and ultimately changing the final answer, a phenomenon referred to as answer flips. To address this instability, we propose FlipDir (Flip-Direction Steering), a training-free inference-time method that estimates a low-rank flip-inducing activation subspace from contrastive pairs of original and answer-flipping inputs and selectively steers hidden states during decoding. A margin-based gate limits subspace attenuation to uncertain decoding steps, recovering original predictions while preserving stable ones. To evaluate robustness beyond accuracy or consistency on fixed test sets, we introduce VisFlip, a
benchmark framework that constructs evaluation groups for a target model and visual variation setting to separately assess recovery of original predictions and preservation of stable ones. VisFlip spans nine
dataset-variation combinations across scientific
reasoning,
robot-scene understanding, and medical VQA, covering subtle visual variations common in each domain. Experiments across 18 settings demonstrate that FlipDir consistently outperforms existing methods on the combined recovery and preservation metric. We will make our code publicly available.
Read the full abstract ▾
All authors Yeonsung Jung, Joonhyun Jeong, Hoang Pham, Joowon Kim, Yoonsik Park, Viet Dac Lai, Eunho Yang
Artificial Intelligence
RECLAIM: Can Agents Reproduce the Claims of Machine Learning Papers?
Mithil Salunkhe, Haochen Ding, Samridhi Verma et al. (+1 more)
Reproducing a machine learning paper involves most research steps, from installing software and debugging to running experiments, work that AI
agents increasingly do. We introduce RECLAIM, a
benchmark of 100 NeurIPS 2025 papers that can be rebuilt yearly from new conferences. For each paper we fix in advance the result to reproduce, what counts as a successful reproduction, and a GPU-hour budget. An
agent must reproduce that result using the paper and whatever its authors released. What the authors released decides the difficulty tier. Run-tier releases include code, data, and weights; Retrain-tier releases lack weights, so the
agent trains the model; Reimplement-tier releases lack code, so the
agent writes it. A separate
language model grades runs from logs and outputs rather than
agents' reports. We run four
agents once per paper; the best
agent in each tier reproduces only 41% of Run-tier papers, 27% at Retrain, and 15% at Reimplement, where every
agent does worst. Failed attempts use on average 29% of their budget, so most stop with budget left. The most common
agent error is writing the method without checking any part against the paper's numbers, in 63 of 400 runs.
Read the full abstract ▾
All authors Mithil Salunkhe, Haochen Ding, Samridhi Verma, Volodymyr Kindratenko
Machine Learning
LastOPD: Taming Collapse in Latent On-Policy Distillation
Jie Yang, Zhengyu Fang, Zelin Xu et al. (+7 more)
On-policy
distillation (OPD) corrects a student on the responses it writes, but its signal is the teacher's next-
token distribution: it tells the student what the teacher says but misses how it thinks. Latent supervision promises the missing part by aligning the student's latent states to the teacher's. Recent methods such as OPRD bring this signal into on-policy
distillation. However, we observe two failures of this recipe when distilling Qwen3-4B and Qwen3-8B into Qwen3-1.7B-Base. Early gain, late collapse: latent supervision alone lifts MATH-500 accuracy from 25 to 46 in 10 steps, but subsequent training degrades performance down to 11 with no recovery. Better
alignment, worse behavior: although the
alignment metric steadily improves throughout this collapse, the most aligned model turns out to be the worst performing. Further analysis suggests a mismatch in how the latent signal is applied: layers paired by depth play different roles in the two models, so continued
alignment may pull the student toward teacher states it cannot understand. To address this, we propose LastOPD, which applies the latent signal only at the last-layer state, the common interface both LM heads read, and only during a 10-step crossfade into
token-level OPD. This keeps the useful part of the latent signal and hands the student to
token-level supervision before the collapse sets in. Extensive experiments show that LastOPD improves MATH-500 over
token-only OPD by 5.55 and 4.02 points with the 4B and 8B teachers, leads on most held-out
datasets, and reaches the final score of
token-only OPD in about half the steps. Code is available at https://github.com/Muyiiiii/LastOPD.
Read the full abstract ▾
All authors Jie Yang, Zhengyu Fang, Zelin Xu, Jiarui Sun, Xiran Fan, Junpeng Wang, Liang Wang, Qinghua Liu, Yiwei Cai, Yan Zheng
Cryptography and Security
Blockchain-Enabled Artificial Intelligence and AI Agents for Secure Data Sharing and Cybersecurity Applications
Harsh Verma
Blockchain and artificial intelligence (AI) are converging into a single infrastructural layer for securing data sharing, model integrity, and autonomous decision-making across distributed systems. This paper presents a meta-synthesis that draws together four constituent studies covering
adversarial machine learning, AI-powered anomaly detection in cloud environments, automated vulnerability patching by multi-
agent large
language model (LLM) pipelines, and the broader landscape of securing AI systems across their lifecycle and situates their findings within the emerging literature on blockchain-enabled AI and autonomous AI
agents. Each constituent study addresses a distinct point of failure in modern AI-driven security operations: the integrity of training data and model behavior, the reliability of real-time monitoring, and the trustworthiness of automated code remediation. We argue that blockchain's properties of immutability, decentralized consensus, and verifiable provenance directly address a gap common to all three: the difficulty of establishing trust in data, models, and autonomous
agents that operate without a central authority. Building on real-world research on blockchain-secured data sharing,
federated learning, and multi-
agent coordination, we propose a layered reference architecture that couples adversarially hardened models, blockchain-anchored data provenance, AI-driven anomaly detection, and smart-contract-governed multi-
agent remediation. We conclude by identifying open problems in scalability,
privacy-transparency trade-offs, and the governance of autonomous
agents that must be resolved before such integrated systems can be trusted in production-critical environments.
Read the full abstract ▾
Computational Complexity
Exponential Correlation Bounds for Polynomials
Eshan Chattopadhyay, Pooya Hatami, Chin Ho Lee et al. (+3 more)
We prove that the XOR of $k$ majorities on disjoint blocks of $\ell$ bits has correlation at most $(4d/\sqrt{\ell})^k$ with every Boolean-valued degree-$d$
polynomial over an arbitrary field. By known techniques, we obtain explicit pseudorandom generators that fool such
polynomials on $n$ Boolean inputs with seed length $d^2\cdot\polylog(n/\varepsilon)$. The correlation and the generator extend to small $\mathsf{AC}^0[q]$
circuits, for every fixed
prime power $q$. These results resolve several long-standing questions in
circuit complexity and pseudorandomness.
Read the full abstract ▾
All authors Eshan Chattopadhyay, Pooya Hatami, Chin Ho Lee, Shachar Lovett, Avishay Tal, Emanuele Viola
Robotics
Uncertainty-Gated Exploration Noise Suppresses Task Collapse in Online RL Fine-Tuning of a Flow-Matching Vision-Language-Action Policy
Mehmet Turan Yard{\i}mc{\i} and Yunus Emre Çoğurcu
Online reinforcement learning fine-tuning of pretrained flow-matching
vision-
language-action (VLA) policies promises
robots that keep learning after deployment, but continued updates often destroy competence on individual tasks while the aggregate still looks healthy. We study this failure mode, which we call task collapse, under a matched small-compute budget on LIBERO-10 with a 450M-parameter SmolVLA policy trained by PPO with
stochastic (SDE) sampling. Three exploration-noise policies differ in one live variable: a fixed noise scale, a ReinFlow-style learned noise network, and an uncertainty-gated controller that redistributes exploration across task streams from task-agnostic novelty and competence signals, without task labels or episode boundaries. Under the pooled definition, fixed noise collapses tasks in two of three seeds and learned noise in every seed measured to iteration 200, while the controller collapses none in any of its three seeds. Measured parameter displacement shows the controller's action expert keeps changing, while its mean applied noise is close to the fixed scale in the available logs. The matched comparison supports the controller's effect on task preservation; the separate contributions of its adaptation across states and over time are not disentangled. A lower fixed scale slows the decline but does not stop it. No arm improves on the behavior-cloning baseline in this budget. Two properties of that regime are measured beside this result, not offered as its cause: following the reference recipe, training runs in bfloat16 with no fp32 master copy, under which 96.02% of the action expert's elements stay bit-identical across three consecutive iterations, and an fp32 master copy at the reference learning rate collapses both arms in a single-seed observation. We release tools measuring per-task collapse under four definitions, rescoring noise and instrument tares.
Read the full abstract ▾
Computer Vision
M$^2$PFN: End-to-End Disentangled Alignment for Generalizable Multimodal In-Context Learning in Alzheimer's Disease
Lujia Zhong, Shuo Huang, Jianwei Zhang et al. (+2 more)
While various
multimodal methods combining imaging and tabular data for Alzheimer's disease (AD) diagnosis were proposed, they are often limited in generalization across cohorts. In-context learning (ICL) has demonstrated excellent generalization performances and high flexibility in foundational tabular models such as TabPFN. To extend TabPFN's ICL to
multimodal AD analysis, the main obstacle is that TabPFN is meta-trained on synthetic tabular priors that do not naturally match the statistical structure of image-derived features. We propose M$^2$PFN, an end-to-end framework that turns this tabular foundation model into a
multimodal AD predictor. M$^2$PFN (i) performs differentiable inference through TabPFN's
transformer, back-propagating task gradients into 3D-MRI and tabular encoders; (ii) aligns the two modalities into a shared subspace, via disentanglement and a contrastive objective, matched to the ICL engine's prior; and (iii) folds in a frozen tabular-only prediction through a learnable gated shortcut. Because the ICL engine stays frozen, its in-context mechanism is preserved for test-time generalization, while end-to-end training shapes the encoders into features it can exploit. On ADNI ($n=2240$, three-class CN/MCI/AD), M$^2$PFN attains $65.55\%$ macro-F1 and $82.21\%$ macro-AUC, surpassing a comprehensive set of unimodal and
multimodal baselines. By swapping only the head for a TabPFN regressor, the same architecture regresses baseline MMSE on a $1250$-subject sub-cohort to test MAE $1.743$, outperforming every
multimodal baseline. On two external cohorts (OASIS-3 and SCAN) with no retraining, M$^2$PFN achieves the best AUC and the lowest MMSE MAE across all baselines, and transfers even when the cognitive instrument changes.
Read the full abstract ▾
All authors Lujia Zhong, Shuo Huang, Jianwei Zhang, Xinyu Nie, Yonggang Shi
Machine Learning
When Does Unsupervised Learning Succeed or Fail? A PoS Perspective on Reconstruction-Based Anomaly Detection
Mehmet Yamaç, Yagmur Mustu, Muhammad Numan Yousaf et al. (+2 more)
Reconstruction-based unsupervised learning can fail in two opposing ways: a model may reconstruct anomalies too accurately or discard valid nominal variation. Using the Pursuit of Subspaces hypothesis, we characterize these failures through the meet, union, and join geometries induced by the nominal components. Excess learned range produces join blindness, while insufficient capacity produces meet preference and loss of nominal fidelity. We show that the compact nominal union is optimal among nominal faithful ranges and generally requires a nonlinear reconstruction map. Based on this geometry, we introduce Dynamic Push and Pull, which learns from controlled perturbations without anomaly labels, and nested
manifold carving, which applies the same principle recursively in latent space. Experiments confirm the predicted changes in latent geometry across every tested Push and Pull configuration. The proposed methods improve reconstruction-based anomaly detection across standard
benchmarks and unseen image degradations, while also improving pretrained ECG representations for downstream classification. These results connect reconstruction failures to identifiable
geometric conditions and provide practical mechanisms for learning compact representations.
Read the full abstract ▾
All authors Mehmet Yamaç, Yagmur Mustu, Muhammad Numan Yousaf, Lei Xu, Marcel van Gerven
Systems and Control
Adaptive State Estimation Under Topological Uncertainty in Unobservable Primary Distribution Systems Using Strategically Placed Sensors
Farah Elsherif, Behrouz Azimian and Anamitra Pal
The rapid integration of distributed energy resources is fundamentally altering power flow patterns in primary distribution networks and intensifying operational uncertainty. These problems are further compounded by lack of real-time situational awareness and frequent
topology changes. To address these problems, this paper proposes an integrated deep learning framework for simultaneous
topology identification (TI) and distribution system state estimation (DSSE) in real-time unobservable primary distribution networks instrumented by a minimal set of synchronized measurement devices (SMDs). A correlation-driven SMD placement
algorithm is introduced first that jointly satisfies TI accuracy and DSSE performance requirements by exploiting temporal and spatial correlations in nodal voltage measurements. A dual deep neural network (DNN)-based DSSE model is developed next to estimate three-phase voltage magnitudes and angles across diverse operating conditions. To extend the framework beyond the base
topology, fine-tuning-based transfer learning is employed to adapt the DSSE model to reconfigured
topologies using limited retraining data. The framework is validated under both Gaussian and non-Gaussian measurement noise and benchmarked against a conventional estimation approach and a single DNN model.
Read the full abstract ▾
Systems and Control
Comparison of Multisine Peak Factor Minimization Algorithms for Aircraft System Identification
Justin J. Matt
Two phase-optimized multisine peak factor minimization
algorithms are presented and evaluated. The first
algorithm minimizes peak factor by iteratively clipping the peaks of generated multisine signals. The second
algorithm optimizes peak factor indirectly through minimization of an approximation of the infinity norm of the multisine.
Algorithm performance was evaluated as a function of different signal properties, including the number of harmonics, harmonic spacing, and number of snow harmonics (extra harmonics included for further reduction of the peak factor). The two
algorithms are compared against results obtained by minimizing peak factor directly using a simplex
algorithm, which has been a common approach when designing phase-optimized multisines for system identification flight tests. Sample results show that the clipping and infinity norm
algorithms produced multisine signals with comparable peak factors that were lower than that of the simplex
algorithm. However, the clipping
algorithm runs an order of magnitude faster than the other two
algorithms, which also makes it practical to repeat the
algorithm multiple times to achieve even lower peak factors.
Read the full abstract ▾
Computation and Language
COILD: An Indic-Centric Parallel Corpus and Benchmark for Machine Translation Across Indian Languages
Kshetrimayum Boynao Singh, Nitin Kumar Mishra, Palash Pratim Dutta et al. (+22 more)
Machine translation (MT) for Indian
languages remains constrained by the limited availability of high-quality, Indic-centric parallel corpora and evaluation
benchmarks. Existing
multilingual resources are largely constructed from English-pivot content and often fail to capture the linguistic diversity, cultural complexity, and domain-specific characteristics of Indian
languages. We present COILD, an Indic-centric parallel corpus comprising over 1.16 million human-translated and human-verified sentence pairs, covering 20 Indian
language pairs across the Indo-Aryan, Dravidian, Tibeto-Burman, and Austro-Asiatic
language families. The corpus is built entirely from original Indian
language sources collected from licensed repositories spanning eight domains with direct real-world applicability. Furthermore, we introduce a domain-centric
benchmark comprising 2,000 expert-verified sentences to enable consistent
multilingual and cross-lingual evaluation across Indian
language pairs. To validate the effectiveness of COILD, we fine-tune two representative
multilingual neural machine translation models, IndicTrans2-Distilled and NLLB-200. Experimental results demonstrate consistent improvements across
language pairs, domains, automatic evaluation metrics, and human evaluation, highlighting the effectiveness of high-quality Indic-centric supervision. COILD provides a valuable training and evaluation resource for advancing
multilingual machine translation and future
multilingual language models for Indian
languages.
Read the full abstract ▾
All authors Kshetrimayum Boynao Singh, Nitin Kumar Mishra, Palash Pratim Dutta, Atai Waris Khan, Aparna Kaushik, Avinash Kumar, Deeksha, Deepak Kumar, Saroj Kumar Jha, Saloka Sengupta, Anansa Roy, Umalatha Kannoth, Saifulla Samar, Meena Sharma, Manpreet Kaur, Jyoti Sharma, Ashwini Vaidya, Muralikrishna SN, Md Shad Akhtar, Poonam Bansal, Amita Dev, Sanasam Ranbir Singh, Samit Bhattacharya, Tanmoy Chakraborty, Asif Ekbal
Systems and Control
Network Design against the Bullwhip Effect in Complex Supply Chains
Lauren Bogo and Dominic Liao-McPherson
This paper studies the bullwhip effect, the amplification of demand fluctuations into larger order fluctuations upstream, in supply chain networks where each firm orders from several suppliers over routes with different lead times. We show that on a directed acyclic network every response from demand to orders is a sum over paths of products of nodal responses, so that the stability and stability margins of the network are determined by those of its individual nodes, and that a node which splits its orders across routes of different lengths gains margin it can spend on a higher gain and a faster response. We then minimize the worst-case amplification of the network by choosing how each firm splits its orders among its suppliers, which route lead times to shorten, and each firm's ordering gain, with gradients from one adjoint pass per demand node and frequency. On a 14-city network the optimized design amplifies 20~dB less than cost-minimal routing and 17~dB less than a design that shortens the longest routes with the same budget, and is better connected than either. Amplification is therefore a property of routing and delay, not only of the ordering policy, and designing against it also makes the network more robust to a lost route.
Read the full abstract ▾
Programming Languages
Pack Iteration in Swift: Ordinary Control Flow for Variadic Generics
Serafima Nerush and Eitan Frachtenberg
Variadic generics are a powerful tool for type-safe meta-programming. Yet in most widely used
languages, they remain an "expert-only" feature due to their reliance on complex patterns such as recursive decomposition or expansion expressions that do not compose naturally with ordinary control flow. In C++, for example, accessing elements of parameter packs has traditionally relied on unintuitive recursive patterns that "peel off" elements. This paper presents Pack Iteration, a feature introduced in Swift 6.0 that allows developers to iterate over parameter packs using a familiar, imperative for-in loop. By treating pack expansion as a first-class source for iteration, Swift bridges the gap between high-level expressiveness and advanced generic programming. We detail the design and implementation of this feature within the Swift compiler, focusing on the challenges of bridging static type-checking in the constraint system with dynamic execution in the Swift Intermediate
Language. Unlike traditional models that expand packs at compile time, Swift's implementation supports on-demand evaluation, enabling efficient dynamic iteration and short-circuiting control flow. Our empirical evaluation confirms that pack iteration provides performance comparable to - and sometimes significantly better than - the complex workarounds previously required.
Read the full abstract ▾
Computers and Society
AI-Enabled Human Memory Manipulation: Misleading AI-Generated Summaries Distort Human Memory
Mattea Sim, Yael Eiger and Tadayoshi Kohno
AI-generated summaries are increasingly used in high-stakes settings, like policing, despite considerable evidence that AI often generates misleading or inaccurate information. This research asked: do errors in AI-generated summaries distort human
memory? To answer this question, we adopted two methodological approaches. First, we conducted an analysis of AI summary output, prompting large
language models to generate summaries of
videos. This analysis quantified how often AI summaries contain errors and the categories of these errors, revealing the kinds of misleading information that may distort human
memory. Second, we conducted a human-subjects experiment to test the impact of misleading information in AI-generated summaries on human
memory. Participants were first exposed to an event via watching a
video of a car-pedestrian accident, and later read an AI-generated summary describing the
video that either contained misleading or accurate information. Participants'
memory for the original event was assessed in a
memory recognition test. In the AI analysis, we found a high frequency of mistakes in AI summaries, and particularly frequent omissions of critical details. For instance, the majority of summaries omitted the most central event of the
video, a critical error which is likely to be impactful. We also observed a strong effect of AI misinformation on human
memory. People who read a misleading AI summary were significantly less likely to accurately recall the original event, compared to people who read an accurate AI summary. These findings have implications for how AI should be used in critical settings. Though "humans-in-the-loop" are often expected to correct for AI's mistakes, our work suggests human
memory can instead be distorted by these mistakes. AI has the potential to generate misinformation, even absent any
adversarial intent, which can meaningfully impact human
memory.
Read the full abstract ▾
Robotics
KeyGen: Unsupervised Keypoint based Object-Centric Representations for Category-Level Policy Generalization
Shuxin Cao, Liquan Wang, Masoud Moghani et al. (+2 more)
Generalization in robotic manipulation requires policies to perform tasks across diverse unseen object instances that vary in shape, size, and pose. However, conventional behavior cloning (BC) methods often overfit to instance-specific geometry and appearance, limiting transfer to novel objects. We introduce KeyGen, a framework that learns canonicalized semantic 3D keypoints from point clouds and uses them as structured object-centric representations for policy learning. A visuomotor
diffusion policy conditions on these keypoints together with object-centric geometry to predict full manipulation trajectories, enabling consistent
geometric correspondence across object instances. To evaluate category-level generalization, we construct a photorealistic simulation
benchmark with three manipulation tasks and a planning-driven data generation pipeline that produces expert trajectories across diverse object instances. Experiments show that KeyGen significantly outperforms prior methods on both seen and unseen objects under pose variation, scales effectively with additional demonstrations per object, maintains robustness to object rescaling, and achieves strong performance in both simulation and real-world manipulation.
Read the full abstract ▾
All authors Shuxin Cao, Liquan Wang, Masoud Moghani, Benjamin Joffe, Animesh Garg
Robotics
FlyCNS: Connectome-Grounded Information Organization for Communication-Constrained Embodied Control
Jinchang Zhang, Jiakai Lin and Guoyu Lu
Robotic bodies are inherently distributed in sensing and actuation, yet learning-based control still commonly relies on centralized information processing. This work studies the problem of information organization in communication-constrained
embodied control: which computations should remain local, and which information is worth transmitting for whole-body coordination. We propose FlyCNS, an
embodied information-organization framework inspired by the Drosophila
brain–nerve-cord connectome. FlyCNS preserves local sensorimotor computation within each limb and enables selective long-range communication through separate ascending and descending routing pathways. From a real connectome, FlyCNS extracts the directional structural complexity of these two pathway types and uses it as a weak prior over communication allocation, while message content, transmission timing, and locomotion policies remain task-adaptive and are learned through reinforcement learning. In Unitree Go1 simulation, FlyCNS exhibits more graceful performance degradation as the communication budget is tightened. Under the most restrictive setting, it uses only about 21–22% of the communication of the full-communication reference, while still maintaining a tracking score of approximately 0.882 under both command protocols, with a gap of no more than 6.1% from the full-communication reference. These results indicate that real neural connectomes can inform not only the structural design of control networks, but also provide transferable inductive biases for information organization across embodiments, guiding
robots in balancing local computation and long-range coordination under limited communication resources.
Read the full abstract ▾
Computer Vision
CinematicVQA: Benchmarking Film-Grammar Reasoning in Large Vision-Language Models
Shuo Xing, Pooja Verlani, Balu Adsumilli et al. (+1 more)
Cinematography, the craft of visual storytelling through framing, lighting, and camera operation, fundamentally shapes how audiences perceive and emotionally engage with
video content. While Large
Vision Language Models (LVLMs) have made remarkable progress in
video question answering, existing
benchmarks primarily focus on identifying low-level techniques rather than understanding their storytelling impact. To address this, we introduce CinematicVQA, the first-of-its-kind
benchmark for cinematic
video understanding that goes beyond technique recognition to evaluate film-grammar
reasoning, utilizing our introduced Cinematic Scene Graph (CSG), a structured representation that links filming techniques to their perceptual effects and narrative functions. Through comprehensive evaluation of state-of-the-art LVLMs, we reveal a striking semantic gap: models consistently perform higher on describing visual presentations than on identifying the underlying techniques. Surprisingly, Chain-of-Thought prompting fails to provide consistent gains and degrades performance for most models, suggesting that current LVLMs lack sufficient cinematic domain knowledge to benefit from step-by-step
reasoning. Fine-tuning on CinematicVQA-train yields consistent improvements, particularly for narrative function and multi-hop
reasoning. Overall, CinematicVQA serves both as a rigorous
benchmark for cinematic evaluation in LVLMs and as a practical
dataset for training more film-aware
video models.
Read the full abstract ▾
All authors Shuo Xing, Pooja Verlani, Balu Adsumilli, Zhengzhong Tu
Human-Computer Interaction
Xiaoran Yang, Yang Zhan, Xie He et al. (+5 more)
LLM-powered conversational
agents (CAs) often present uncertainty and provenance cues alongside their responses to help users assess response reliability and identify potential
hallucinations. In immersive environments such as Virtual Reality (VR), CAs often take the form of
speech-based
embodied conversational
agents (ECAs), where uncertainty and provenance cues cannot rely on persistent inline text and may be missed or disrupt comprehension when delivered through
speech. We conducted a within-subjects study (N = 24) to compare three designs for presenting the
hallucination-awareness information (uncertainty and provenance) in ECAs in VR against a no-cue baseline:
embodied cues using gestures and posture, icon cues using visual indicators, and text cues using color-coded text with inline citations. We evaluated how these designs affect users' ability to identify
hallucination-related information, trust in the ECA, and interaction experience (immersion and task load). Our results show that all three designs support users in identifying
hallucinations.
Embodied cues were associated with higher trust and immersion, text cues offered clearer interpretability, and icon cues preserved relatively good interpretability while causing less disruption to immersion compared with
embodied cues and text cues. This work contributes to the VR and AI research community by comparing different designs of
hallucination cues in immersive ECA settings and examining how they affect users' ability and experiences to identify
hallucinations. It also offers practical insights and design implications for developing future
hallucination-awareness interfaces for ECA.
Read the full abstract ▾
All authors Xiaoran Yang, Yang Zhan, Xie He, Yuxuan Huang, Yichen Yu, Zhuo Wang, Noboru Matsuda, Qiao Jin
Computer Vision
DeltaWAM: Delta World Action Models for Bimanual Manipulation
Han Yan, Zishang Xiang, Haokai Jiang et al. (+6 more)
World-action models (WAMs) transfer visual and motion priors from pretrained
video generators to
robot control by jointly modeling visual dynamics and actions. Existing WAMs, however, predict dense future frames during training, repeatedly modeling largely unchanged content and coupling action-conditioned dynamics to nuisance appearance variations. At inference, processing each complete observation with the heavy
video expert bottlenecks few-step action generation. Accordingly, we propose DeltaWAM, which jointly predicts visual deltas and actions using dense-anchor, sparse-delta, and action streams, with three architectures that differ in representation and computation sharing. We further develop Streaming Delta
Memory (SDM), which updates cached anchor context with compact observed deltas, reducing heavy
video-expert processing. On RoboTwin, DeltaWAM with SDM improves average success over Fast-WAM from 81.3% to 85.4% in the clean setting and from 75.8% to 83.9% under visual randomization. The three architectures reduce training FLOPs by 17.78-23.77%, while SDM reduces one-step inference latency and FLOPs by 36.57% and 31.55%, respectively; real-world evaluations further show the highest overall success rate and normalized progress among the evaluated policies. Code: https://github.com/AIGeeksGroup/DeltaWAM. Website: https://aigeeksgroup.github.io/DeltaWAM.
Read the full abstract ▾
All authors Han Yan, Zishang Xiang, Haokai Jiang, Zeyu Zhang, Qilin Wang, Weiyu Guo, Yandong Guo, Boxin Shi, Hao Tang
Software Engineering
Category-Based MLM: Unifying Powertypes with Superclasses
Mira Balaban, Azzam Maraee and Arnon Sturm
MultiLevel software Modeling (MLM) suggests that conceptual modeling in broad subject domains might require abstraction of multiple classification levels. The MLM approach relies on philosophical arguments, claiming that faithful modeling of real-world domains involves repeated type classification as in ontologies of natural kinds. MLM leveled architecture is interwoven and defined by instance-of interlevel relationships between clabject classes in lower levels to classes termed category classes, in upper levels. The instance-of relation denotes membership of clabjects as type objects in their (powertypes) category classes, and is not transitive. All MLM approaches support forms of deep characterization, i.e., category classes can influence classes in lower levels. Deep characterization is an essential feature of superclasses and contradicts the non-transitive membership meaning of instance-of. In this paper, we introduce the Category-Based MLM (CatMLM) model, in which category classes have dual superclass and powertype facets, based on the distinction between category features that do not participate in deep characterization, and object features that do. This distinction clarifies the role of levels, provides a clear quantifiable criterion for leveling, and yields a decision rule between the subclass and instance-of relations. The contribution of this paper is to introduce a well-defined MLM model that (1) is based on simple, quantifiable level decisions; (2) clarifies how leveling emerges from domain needs; and (3) analyzes gains and losses of MLM vs. plain OO modeling.
Read the full abstract ▾
Machine Learning
Stream Recursion Model (SRM)
Asael Sorensen, Charles Brock, David Chamberlain et al. (+3 more)
Mechanistic interpretability seeks to make verifiable statements about the internal behavior of large
language models (LLMs). Many interpretability techniques struggle to scale with the increasing size and depth of architectures. Our solution to this is to introduce smaller models with structures that lend themselves to interpretability. In this work, we introduce the Stream Recursion Model (SRM), a modification of the Hierarchical
Reasoning Model (HRM) designed to expose internal computational structure while remaining scalable. SRM organizes computation into multiple interacting latent streams that are updated through recursive refinement, enabling direct analysis of stream dynamics, causal contribution, and routing behavior. SRM achieves performance comparable to GPT-2 on a per-parameter basis. Our analysis reveals consistent and distinct behavior across streams, indicating structured specialization and interaction. These results suggest that SRM provides a practical architectural foundation for scalable mechanistic interpretability and opens up promising avenues for future research in both
reasoning performance and interpretability.
Read the full abstract ▾
All authors Asael Sorensen, Charles Brock, David Chamberlain, Jennifer Minnich, Matthew Hoffman, Ramyaa Ramyaa
Robotics
An Analysis of Streaming Deep Reinforcement Learning for Adaptive Continual Learning in Robotics
Teeratham Vitchutripop, Alyssa Quarles, Wenhe Zhang et al. (+2 more)
Over the course of a lifetime,
robots may encounter novel scenarios unaccounted for in its original training that result in performance degradation. One common approach to mitigating this issue is to further grow the offline training
dataset in hopes of producing a policy robust to these changes. In contrast, biological learning occurs moment-to-moment via a stream of experience, unlike the predominantly batch-based and offline nature of deep learning. Although recent works show the feasibility of stream-based deep reinforcement learning, where updates use only the latest experience, none have shown it to be a viable continual learning framework for adapting robotic policies to unseen changes. In this paper, we present the first analysis of streaming deep reinforcement learning for adaptive continual learning in robotics. In particular, we show that, following an initial pretraining phase, streaming deep RL can enable a
robot to successfully adapt to unforeseen changes to itself, its environment, or goals. Our primary experiments within quadruped locomotion demonstrate that a deep neural network robotic policy with certain optimizers and plasticity loss mitigation techniques can successfully leverage domain task knowledge from its pretraining to quickly adapt online to diverse changes via stream learning, outperforming batch-based on-policy methods and improving task success rates by up to 90% over the pretrained policy. Furthermore, we perform additional evaluations on robotic manipulation tasks to determine if our previous observations extend to different robotic morphologies and scenarios. Our results show that the successes observed in quadruped locomotion can be partially realized in manipulation with stability and performance limitations. We conclude with a discussion on the limitations of our work and its implications for the future of continual
robot learning.
Read the full abstract ▾
All authors Teeratham Vitchutripop, Alyssa Quarles, Wenhe Zhang, Richard Xue, Daniel Rakita
Artificial Intelligence · Cross-listed from eess.AS
A Harness for Synthesizing Diverse Naturalistic Full-Duplex Conversations
Matthew Sun, Vinay Kothapally, Meng Yu et al. (+4 more)
Full-duplex dialogue systems, which listen while speaking, must distinguish a completed turn from a pause within a turn and an interruption that requests a turn from a brief acknowledgment or
speech addressed to a third party. Yet existing conversational corpora provide limited control over these events and limited labels for their intent. We present a pipeline for synthesizing intent-labeled, two-channel conversational
speech from relational event lists. An LLM authors each event's speaker, text, conversational act, and attachment to an earlier event without predicting absolute timestamps. Events are synthesized independently, aligned with their source text, and placed on a shared clock, so turn-taking landmarks are measured from the rendered signal while silence durations are specified or sampled from turn-taking distributions. The pipeline covers 42 phenomena across eight families in English and Mandarin, derives frame-level system actions from authored intent, and promotes diversity using small, diverse sets of prior examples and batch
prompts that request alternatives with self-reported probabilities. Ablations show gains in each targeted diversity dimension. On a four-action label space for taking, holding, releasing, and not holding the conversational floor, a semantic voice-activity detector using only current and past audio reaches start-speaking and start-listening F1 scores of 0.819 and 0.802. When generating its own responses, the full-duplex
speech model Moshi takes 0.85 of the reference turns after fine-tuning on the generated corpus, compared with 0.44 before fine-tuning. Its frame-level precision for predicting system-floor occupancy rises from 0.46 to 0.88. With reference context at each step, its frame-level floor F1 rises from 0.893 to 0.962. These results show that controlled synthesis can provide learnable and transferable supervision for full-duplex turn management.
Read the full abstract ▾
All authors Matthew Sun, Vinay Kothapally, Meng Yu, Chao Huang, Hao Zhang, Yixuan Zhang, Steve Yves
Human-Computer Interaction
The Interface Is Downstream: Designing the Terms of Human-Agent Collaboration
Hector Ouilhet Olmos
Before an
agent responds or acts, much of the experience has already been designed.
Memory and
retrieval shape what it notices. Evidence rules shape what it may claim. Permissions shape what it can do. Learning rules shape what it carries into the next encounter. The argument comes from Alicia, a personal
agent I've built and used since January 2026. A fine-tuning pilot produced no defensible model-performance result. It exposed a provenance failure: Alicia repeated an interpretation from a retrieved synthesis, cited a source note credited by that synthesis, and left the synthesis out of the visible chain. In a model-blind review of thirty-three citations, one reviewer judged that the retrieved intermediary supplied the claim in sixteen relayed citations and part of it in four. Five of twelve citations to directly retrieved targets lacked support in the target excerpt supplied for review. These judgments remain unadjudicated, and the packet is not public. I call the shared setting a humorphic environment: a persistent computational setting that translates a human practice into software. The first Humorphism paper translated partnership. This paper translates the studio, the room where practice happens. The failure prompted an audit of
attention, evidence, action, and learning, with a review artifact and available recourse for each. The test is whether the person can inspect and contest what shaped the teammate's behavior. The output is downstream. Correction, consent, and learning carry the collaborative interface back upstream.
Read the full abstract ▾
Numerical Analysis
Two-Dimensional Shallow Water Linearized Moment Equations: Hyperbolicity and Well-Balanced Schemes
Shiping Zhou, Juntao Huang and Andrew Christlieb
We study rotational invariance, hyperbolicity, and well-balanced discretization of shallow water linearized moment equations in two horizontal dimensions. Although the direct extension has real characteristic speeds, its directional
matrices can lack a complete eigenbasis. We construct a rotationally invariant modification and prove its global hyperbolicity for all states with positive water depth. For quasi-two-dimensional flows over non-flat topography with Navier-slip friction, we develop first- and second-order path-conservative finite-volume schemes that exactly preserve a discrete moving
equilibrium constructed by midpoint collocation. Numerical tests confirm
equilibrium preservation to roundoff, second-order approximation of smooth stationary branches, and near-second-order accuracy of the second-order scheme for the tested perturbations. Comparisons with three-dimensional two-phase OpenFOAM simulations illustrate the invariant shallow-water limit and show that retaining moments improves the reconstruction of curved rotational velocity profiles, while differences remain in radial motion.
Read the full abstract ▾
Robotics
OCC4M: Object-Centric 4D Memory for Spatiotemporal Reasoning in Long-Horizon Manipulation
Jack B. Jedlicki, Tanguy Dieudonné and Heng Yang
Long-horizon manipulation often requires
reasoning about state absent from the current view, such as a vanished object's location, temporal identity, or the contents of a shuffled container. We present OCC4M ("Occam"), an object-centric 4D
memory that maintains persistent tracks in a shared world frame and explicitly represents temporal, motion, and containment relations. A
vision-
language model (VLM) queries this structured
memory to select actionable targets for history-free low-level execution. Across seven simulation conditions and 350 episodes, OCC4M achieves 96.6%
memory success and 88.9% end-to-end success, versus 54.6% and 57.7% for FrameSamp, a raw-history VLM baseline using Gemini 3.7 Flash with the complete observation history and the same executor. In a controlled viewpoint-transfer test, OCC4M maintains 100%
memory and 98% end-to-end success after a viewpoint change, while full-history FrameSamp falls to near-zero success. On 20 fixed-camera Franka episodes, OCC4M reaches 85% joint
memory accuracy, versus at most 30% for FrameSamp across context sizes from $K=16$ to the complete history, and completes 45% of full two-stage tasks. These results support explicit object-centric
memory for persistent spatiotemporal
reasoning in long-horizon manipulation. Qualitative
videos are available at https://occ4m-sup.github.io/occ4m-supplementary/.
Read the full abstract ▾
Computer Vision
DrGait: Biomechanically Grounded Visual Reasoning for Interpretable Clinical Gait Analysis
Xiangyu Yin, Shiqi Wang, Abrar Alamri et al. (+4 more)
Current automated gait analysis for clinical applications relies on uninterpretable black-box classifiers. Although
Vision-
Language Models (VLMs) offer strong
reasoning capabilities, applying them directly to gait
videos often leads to
hallucinations, because they struggle to measure subtle
geometric deviations from raw visual contexts. To address this, we introduce DrGait, a training-free agentic framework that shifts the VLM's role from a direct visual reasoner to a clinical planner. DrGait decouples semantic
reasoning from
geometric perception through a structured Triage-Verification-Synthesis (TVS) workflow. Given an input
video and a set of basic spatiotemporal metrics, the DrGait
agent first performs a heuristic triage to propose diagnostic hypotheses, which are then verified by autonomously calling deterministic biomechanical tools that operate on reconstructed 3D mesh trajectories, segmented 2D pose tracks, and event-centered
video evidence. Finally, a closed-loop mechanism recursively updates the
agent's
reasoning context based on the feedback. By anchoring VLM's
reasoning in verifiable
geometric and temporal measurements, DrGait reduces
hallucinations, achieving competitive diagnostic accuracy while generating transparent and audit-ready clinical reports.
Read the full abstract ▾
All authors Xiangyu Yin, Shiqi Wang, Abrar Alamri, Yasir Aljohani, Weichen Liu, Goeran Fiedler, Wei Gao