arXiv subjects

Jian-Jia Chen

Publications and source records attributed to Jian-Jia Chen.

At least 19 recordsLinked to original sources

arXiv2026

MISHAP-Bench: A Hallucination Benchmark for Large Audio-Language Models

Large audio-language models (LALMs) produce fluent responses about audio but often hallucinate by making plausible yet ungrounded claims. Existing audio hallucination benchmarks mainly measure response correctness, leaving it unclear whether an LALM hallucinates or simply fails to understand the audio. We challenge correctness-based evaluation by defining two hallucination categories: (i) context, where claims are not grounded in the audio; and (ii) knowledge, where claims about audio-related topics lack support from externally verifiable facts. We introduce MISHAP-Bench, a comprehensive benchmark with 12,000 challenging open-ended question-audio pairs and a rigorous evaluation pipeline covering both categories. To evaluate open-ended responses, we propose a groundedness judge that uses reference rubrics and judge prompts guided by human annotations. We extensively evaluate ten state-of-the-art LALMs and show that hallucination remains substantial. Even a frontier model such as Gemini 3.7 Flash reaches a hallucination rate of 36.5%. We further adapt and benchmark four mitigation methods from multiple domains for LALMs. Despite some improvements, effective hallucination mitigation remains an open challenge. Finally, we call on the community to evaluate hallucination and benchmark mitigation methods with MISHAP-Bench.

cs.LG↗

arXiv2026

Diagnosing Temporal Misalignment in Multichannel Time-Series Classification with Minimum Description Length

Multichannel time-series classification commonly assumes synchronized sensor streams, although latency, clock drift, and preprocessing can introduce relative delays during data collection or after deployment. Existing synchronization solutions are often hardware-specific and difficult to apply retrospectively. Consequently, synchronization problems may remain undetected while classification performance is suboptimal. We introduce a classifier- and label-free diagnostic based on minimum description length (MDL). Our method applies candidate temporal shifts to sensor groups and measures how efficiently one group can be encoded through a representation of the remaining channels. An increased codelength indicates that the shift destroys shared temporal structure, whereas the minimum identifies the alignment most strongly supported by the data. Unlike learned synchronization methods, the diagnostic requires neither retraining nor a trusted aligned reference and can therefore test both training and deployment data for misalignments. Experiments on two controlled synthetic tasks and nine real-world datasets show that the metric exposes alignment structure and can recover accuracy under induced deployment drift. A whole-dataset audit further identifies stable nonzero MDL optima in established benchmarks including FordChallenge, Opportunity, PAMAP2, and UCIActivity, revealing potential systematic offsets that conventional model evaluation does not expose. Our method thus provides a general-purpose tool for detecting, understanding, and correcting temporal misalignment throughout the time-series learning pipeline. Our code is available under https://github.com/sbuschjaeger/mdl-temporal-misalignment.

cs.LG↗

arXiv2026

Deductive Verification for Earliest Deadline First Scheduler Implementations

Real-Time Operating Systems (RTOSes) rely on scheduler implementations to provide predictable task execution. For safety-critical systems, it is therefore not sufficient to reason only about the abstract scheduling policy; the concrete implementation must also preserve the intended scheduling semantics. This is particularly challenging for Earliest Deadline First (EDF) scheduling, because EDF introduces dynamic, deadline-derived priorities that are often realized by reusing kernel infrastructure originally designed for fixed-priority scheduling. In this work, we formalize EDF correctness through three essential properties that any implementation of the Earliest Deadline First (EDF) scheduler must satisfy. Based on these properties, we propose a framework utilizing deductive verification, that applies to any EDF-based scheduler realization. We instantiate the framework in Frama-C/ACSL and apply it to three structurally different EDF scheduler realizations: RTEMS 5, RTEMS 6, and an EDF extension of FreeRTOS.

cs.OS↗

arXiv2026

Schedulable Job-Level Dependencies for Cause-Effect Chains via Graph Neural Networks

Modern automotive software architectures comprise large sets of mixed-criticality functions executing on shared multi-core platforms with strict real-time and end-to-end timing requirements. Sensor-to-actuator data propagation in such systems is typically expressed via cause-effect chains with worst-case data-age budgets. Job-level dependencies (JLDs) have been introduced to provide a schedule-agnostic mechanism for bounding the data age independently of the underlying scheduler. The state-of-the-art methods for synthesizing JLDs, however, do not check whether the produced JLDs are enforceable under a concrete scheduling policy or jointly schedulable at the system level. In this paper we propose the first machine-learning-based JLD synthesis method, built around a two-level Graph Neural Network with temperature-controlled sampling that learns the structural patterns connecting cause-effect chain configurations to their JLD solutions. Since learned outputs may not be correct by construction, we embed the GNN in a novel Generate-and-Verify architecture in which a safe DP data-age checker, together with a per-chain EDF feasibility checker and a system-level demand-bound test, accept or reject each candidate. We show that the ML-based generator substantially outperforms the original greedy heuristic while achieving orders-of-magnitude lower synthesis time, demonstrating that learned structural priors can effectively replace exponential propagation-tree enumeration on this class of real-time scheduling problems.

cs.SE↗

arXiv2026

Dustin: Draft-Augmented Sparse Verification for Efficient Long-Context Generation with Speculative Decoding

While speculative decoding improves inference throughput for multi-batch long-context Large Language Models (LLMs), its efficiency is often limited by a verification bottleneck where Key-Value (KV) cache loading dominates latency. Existing compression methods fail in this regime: static eviction incurs accuracy loss due to saliency shift, while dynamic selection introduces prohibitive computational overhead during the verification path. We propose Dustin, a sparse verification framework designed for long-context speculative decoding. Dustin integrates lookahead signals from the draft model with historical attention from the target model to identify critical tokens with high fidelity across multi-step verification windows. To reduce recomputation latency, this approach further employs a sparse estimation scheme that restricts importance scoring to a minimal subset of attention heads. Evaluations on PG-19 and LongBench with Qwen2.5-72B demonstrate that Dustin achieves a 27.85x speedup in self-attention and a 9.17x end-to-end decoding speedup at a 32k sequence length, all with negligible accuracy degradation.

cs.CL↗

arXiv2026

Deterministic Execution of ROS 2 Applications via Lingua Franca

The Robot Operating System 2 (ROS 2) is a widely used middleware for robotic systems, characterized by a publish-subscribe (pub-sub) communication mechanism in which computation is structured as callbacks dispatched by ROS 2 executors. Despite its popularity, the pub-sub pattern in ROS 2 is inherently nondeterministic: the order in which these callbacks run is nondeterministic even within a single executor, and distributed deployments add further nondeterminism from the interleaving of messages across nodes and from network latency. Such nondeterminism often leads to concurrency issues and makes it virtually impossible to analyze for safeness and provide guarantees. We present a framework that is able to convert an unmodified ROS 2 application and run it under Lingua Franca (LF), a coordination language for deterministic execution using logical time, so that the same input always produces the same deterministic execution order. We first describe which ROS 2 features can be executed deterministically under logical time. Such features enable the possibility to establish an automatic conversion framework to extract information from a ROS 2 application and directly convert it into an LF program. The rich features of LF, such as logical-time delays, federated execution across processes, and fault handling, can then be applied to make the ROS 2 application be executed in a deterministic and timing-predictable manner without changing the ROS 2 code. We evaluate the framework on a synthetic example and on the Autoware reference system. We show that the order in which callbacks are executed differs in default ROS 2, while also having end-to-end latencies that vary across executions. In contrast, our LF-controlled ROS 2 system produces a deterministic execution order and consistent end-to-end latencies.

cs.RO↗

arXiv2026

Detecting Diffusion-Generated Time Series Under Generator Shift

The boundary between real and diffusion-generated time series is becoming increasingly difficult to draw, yet detection in this domain remains underexplored, especially when the generator is unknown. We compare white-box detection, which requires access to the generator, against black-box detection, which operates on the raw signal alone. The white-box approach, a reconstruction-based detector adapted from the image domain, works well in in-distribution but breaks down under generator shift: reconstruction-based detection in images succeeds because large generic generators provide a near-universal reconstruction prior, and no analogous generator exists for time series. In contrast, a simple off-the-shelf classifier used as a black-box detector performs remarkably well, achieving an average F1 of 79.2, a 22.1% relative improvement over the white-box approach, and a TPR@1%FPR of 57.2. Diffusion-generated time series detection is therefore not a direct transfer of the image domain problem. This work provides the first systematic exploration of white-box and black-box detection for diffusion-generated time series. We close by identifying several open and promising directions.

cs.LG↗

arXiv2026

A Survey of Real-Time Support, Analysis, and Advancements in ROS 2

The Robot Operating System 2 (ROS~2) has emerged as a relevant middleware framework for robotic applications, offering modularity, distributed execution, and communication. In the last six years, ROS~2 has drawn increasing attention from the real-time systems community and industry. This survey presents a comprehensive overview of research efforts that analyze, enhance, and extend ROS~2 to support real-time execution. We first provide a detailed description of the internal scheduling mechanisms of ROS~2 and its layered architecture, including the interaction with DDS-based communication and other communication middleware. We then review key contributions from the literature, covering timing analysis for both single- and multi-threaded executors, metrics such as response time, reaction time, and data age, and different communication modes. The survey also discusses community-driven enhancements to the ROS~2 runtime, including new executor algorithm designs, real-time GPU management, and microcontroller support via micro-ROS. Furthermore, we summarize techniques for bounding DDS communication delays, message filters, and profiling tools that have been developed to support analysis and experimentation. To help systematize this growing body of work, we introduce taxonomies that classify the surveyed contributions based on different criteria. This survey aims to guide both researchers and practitioners in understanding and improving the real-time capabilities of ROS~2.

cs.RO↗

arXiv2026

TMPDiff: Temporal Mixed-Precision for Diffusion Models

Diffusion models are the go-to method for Text-to-Image generation, but their iterative denoising processes has high inference latency. Quantization reduces compute time by using lower bitwidths, but applies a fixed precision across all denoising timesteps, leaving an entire optimization axis unexplored. We propose TMPDiff, a temporal mixed-precision framework for diffusion models that assigns different numeric precision to different denoising timesteps. We hypothesize that quantization errors accumulate additively across timesteps, which we then validate experimentally. Based on our observations, we develop an adaptive bisectioning-based algorithm, which assigns per-step precisions with linear evaluation complexity, reducing an otherwise exponential search problem. Across four state-of-the-art diffusion models and three datasets, TMPDiff consistently outperforms uniform-precision baselines at matched speedup, achieving 10 to 20% improvement in perceptual quality. On FLUX.1-dev, TMPDiff achieves 90% SSIM relative to the full-precision model at a speedup of 2.5x over 16-bit inference.

cs.CV↗

arXiv2026

Jump Like A Squirrel: Optimized Execution Step Order for Anytime Random Forest Inference

Due to their efficiency and small size, decision trees and random forests are popular machine learning models used for classification on resource-constrained systems. In such systems, the available execution time for inference in a random forest might not be sufficient for a complete model execution. Ideally, the already gained prediction confidence should be retained. An anytime algorithm is designed to be able to be aborted anytime, while giving a result with an increasing quality over time. Previous approaches have realized random forests as anytime algorithms on the granularity of trees, stopping after some but not all trees of a forest have been executed. However, due to the way decision trees subdivide the sample space in every step, an increase in prediction quality is achieved with every additional step in one tree. In this paper, we realize decision trees and random forest as anytime algorithms on the granularity of single steps in trees. This approach opens a design space to define the step order in a forest, which has the potential to optimize the mean accuracy. We propose the Optimal Order, which finds a step order with a maximal mean accuracy in exponential runtime and the polynomial runtime heuristics Forward Squirrel Order and Backward Squirrel Order, which greedily maximize the accuracy for each additional step taken down and up the trees, respectively. Our evaluation shows, that the Backward Squirrel Order performs $\sim94\%$ as well as the Optimal Order and $\sim99\%$ as well as all other step orders.

cs.LG↗

arXiv2025

Fixed-Priority and EDF Schedules for ROS2 Graphs on Uniprocessor

This paper addresses limitations of current scheduling methods in the Robot Operating System (ROS)2, focusing on scheduling tasks beyond simple chains and analyzing arbitrary Directed Acyclic Graphs (DAGs). While previous research has focused mostly on chain-based scheduling with ad-hoc response time analyses, we propose a novel approach using the events executor to implement fixed-job-level-priority schedulers for arbitrary ROS2 graphs on uniprocessor systems. We demonstrate that ROS 2 applications can be abstracted as forests of trees, enabling the mapping of ROS 2 applications to traditional real-time DAG task models. Our usage of the events executor requires a special implementation of the events queue and a communication middleware that supports LIFO-ordered message delivery, features not yet standard in ROS2. We show that our implementation generates the same schedules as a conventional fixed-priority DAG task scheduler, in spite of lacking access to the precedence information that usually is required. This further closes the gap between established real-time systems theory and ROS2 scheduling analyses.

cs.DC↗

arXiv2025

On the Limitations of Language Targeted Pruning: Investigating the Calibration Language Impact in Multilingual LLM Pruning

Recent advances in large language model (LLM) pruning have shown state-of-the-art (SotA) compression results in post-training and retraining-free settings while maintaining high predictive performance. However, previous research mainly considered calibrating based on English text, despite the multilingual nature of modern LLMs and their frequent use in non-English languages. This analysis paper conducts an in-depth investigation of the performance and internal representation changes associated with pruning multilingual language models for monolingual applications. We present the first comprehensive empirical study, comparing different calibration languages for pruning multilingual models across diverse languages, tasks, models, and SotA pruning techniques. We further analyze the latent subspaces, pruning masks, and individual neurons within pruned models. Our results reveal that while calibration on the target language effectively retains perplexity and yields high signal-to-noise ratios, it does not consistently improve downstream task performance. Further analysis of internal representations at three different levels highlights broader limitations of current pruning approaches: While they effectively preserve dominant information like language-specific features, this is insufficient to counteract the loss of nuanced, language-agnostic features that are crucial for knowledge retention and reasoning.

cs.CL↗

arXiv2025

CertMask: Certifiable Defense Against Adversarial Patches via Theoretically Optimal Mask Coverage

Adversarial patch attacks inject localized perturbations into images to mislead deep vision models. These attacks can be physically deployed, posing serious risks to real-world applications. In this paper, we propose CertMask, a certifiably robust defense that constructs a provably sufficient set of binary masks to neutralize patch effects with strong theoretical guarantees. While the state-of-the-art approach (PatchCleanser) requires two rounds of masking and incurs $O(n^2)$ inference cost, CertMask performs only a single round of masking with $O(n)$ time complexity, where $n$ is the cardinality of the mask set to cover an input image. Our proposed mask set is computed using a mathematically rigorous coverage strategy that ensures each possible patch location is covered at least $k$ times, providing both efficiency and robustness. We offer a theoretical analysis of the coverage condition and prove its sufficiency for certification. Experiments on ImageNet, ImageNette, and CIFAR-10 show that CertMask improves certified robust accuracy by up to +13.4\% over PatchCleanser, while maintaining clean accuracy nearly identical to the vanilla model.

cs.CV↗

arXiv2025

Speculate Deep and Accurate: Lossless and Training-Free Acceleration for Offloaded LLMs via Substitute Speculative Decoding

The immense model sizes of large language models (LLMs) challenge deployment on memory-limited consumer GPUs. Although model compression and parameter offloading are common strategies to address memory limitations, compression can degrade quality, and offloading maintains quality but suffers from slow inference. Speculative decoding presents a promising avenue to accelerate parameter offloading, utilizing a fast draft model to propose multiple draft tokens, which are then verified by the target LLM in parallel with a single forward pass. This method reduces the time-consuming data transfers in forward passes that involve offloaded weight transfers. Existing methods often rely on pretrained weights of the same family, but require additional training to align with custom-trained models. Moreover, approaches that involve draft model training usually yield only modest speedups. This limitation arises from insufficient alignment with the target model, preventing higher token acceptance lengths. To address these challenges and achieve greater speedups, we propose SubSpec, a plug-and-play method to accelerate parameter offloading that is lossless and training-free. SubSpec constructs a highly aligned draft model by generating low-bit quantized substitute layers from offloaded target LLM portions. Additionally, our method shares the remaining GPU-resident layers and the KV-Cache, further reducing memory overhead and enhance alignment. SubSpec achieves a high average acceptance length, delivering 9.1x speedup for Qwen2.5 7B on MT-Bench (8GB VRAM limit) and an average of 12.5x speedup for Qwen2.5 32B on popular generation benchmarks (24GB VRAM limit).

cs.CL↗

arXiv2025

Optimizing Age-of-Information in Piggyback Networks with Recurrent Data Generation

Age-of-information (AoI) is a critical metric that quantifies the freshness of data in communication systems. In the era of the Internet of Things (IoT), data collected by resource-constrained devices often need to be transmitted to a central server to extract valuable insights in a timely manner. However, maintaining a stable and direct connection between a vast number of IoT devices and servers is often impractical. The Store-Carry-Forward (SCF) communication paradigm, such as Piggyback networks, offers a viable solution to address the data collection and transmission challenges in distributed IoT systems by leveraging the mobility of mobile nodes. In this work, we investigate AoI within the context of patrolling data collection drones, where data packets are generated recurrently at devices and collected by a patrolling drone to be delivered to a server. Our objective is to design a patrolling route that minimizes the Maximum Age-of-Information (MAI) across the system. We demonstrate that determining whether a route with an MAI below a certain threshold can be constructed is NP-Complete. To address this challenge, we propose two approaches with approximation guarantees. Our evaluation results show that the proposed approaches can achieve near-optimal routes in reasonable time across various scenarios

cs.DS↗

arXiv2025

You Only Look Once at Anytime (AnytimeYOLO): Analysis and Optimization of Early-Exits for Object-Detection

We introduce AnytimeYOLO, a family of variants of the YOLO architecture that enables anytime object detection. Our AnytimeYOLO networks allow for interruptible inference, i.e., they provide a prediction at any point in time, a property desirable for safety-critical real-time applications. We present structured explorations to modify the YOLO architecture, enabling early termination to obtain intermediate results. We focus on providing fine-grained control through high granularity of available termination points. First, we formalize Anytime Models as a special class of prediction models that offer anytime predictions. Then, we discuss a novel transposed variant of the YOLO architecture, that changes the architecture to enable better early predictions and greater freedom for the order of processing stages. Finally, we propose two optimization algorithms that, given an anytime model, can be used to determine the optimal exit execution order and the optimal subset of early-exits to select for deployment in low-resource environments. We evaluate the anytime performance and trade-offs of design choices, proposing a new anytime quality metric for this purpose. In particular, we also discuss key challenges for anytime inference that currently make its deployment costly.

cs.CV↗

arXiv2025

Modeling and Simulating Emerging Memory Technologies: A Tutorial

Non-volatile Memory (NVM) technologies present a promising alternative to traditional volatile memories such as SRAM and DRAM. Due to the limited availability of real NVM devices, simulators play a crucial role in architectural exploration and hardware-software co-design. This tutorial presents a simulation toolchain through four detailed case studies, showcasing its applicability to various domains of system design, including hybrid main-memory and cache, compute-in-memory, and wear-leveling design. These case studies provide the reader with practical insights on customizing the toolchain for their specific research needs. The source code is open-sourced.

cs.AR↗

arXiv2025

CHRONOS: Compensating Hardware Related Overheads with Native Multi Timer Support for Real-Time Operating Systems

The management of timing constraints in a real-time operating system (RTOS) is usually realized through a global tick counter. This counter acts as the foundational time unit for all tasks in the systems. In order to establish a connection between a tick and an amount of elapsed time in the real world, often this tick counter is periodically incremented by a hardware timer. At a fixed interval, this timer generates an interrupt that increments the counter. In an RTOS, jobs can only become ready upon a timer tick. That means, during a tick interrupt, the tick counter will be incremented, jobs will be released, and potentially, a scheduling decision will be conducted to select a new job to be run. As this process naturally uses some processing time, it is beneficial regarding the system utilization to minimize the time spent in tick interrupts. In modern microcontrollers, multiple hardware timers are often available. To utilize multiple timers to reduce the overhead caused by tick interrupts, multiple methods are introduced in this paper. The number of interrupts that are triggered by these timers can then be reduced by mapping tasks to timers in such a manner that the greatest common divisor (GCD) of all task periods in a subset is maximized, and the GCD is adopted as the interrupt interval of the timer. To find an optimal mapping of tasks to timers, an MIQCP-model is presented that minimizes the overall number of tick interrupts that occur in a system, while ensuring a correct task release behavior. The presented methods are implemented in FreeRTOS and evaluated on an embedded system. The evaluation of the methods show, that compared to the baseline implementation in FreeRTOS that uses a single timer with a fixed period, the presented methods can provide a significant reduction in overhead of up to $\approx10\times$ in peak and up to $\approx 6\times$ in average.

cs.OS↗