arXiv subjects

Jing Peng

Publications and source records attributed to Jing Peng.

At least 19 recordsLinked to original sources

arXiv2026

Discrete vs. Continuous: A Comprehensive Study of Unified Audio Understanding in LALMs

Large Audio Language Models (LALMs) utilize either continuous features or discrete tokens, yet the optimal representation paradigm for general audio understanding remains debated. Existing benchmarks often focus on narrow domains or evaluate encoders outside LALM contexts. To address these gaps, we systematically evaluate continuous and discrete representations across speech, sound and music. Utilizing our UniARC framework with dual evaluation strategies across model scales from SmolLM2-135M to Llama-3-8B, we analyze the dynamic relationships of data volume, model capacity, and computational efficiency. Our results reveal the pivotal role of semantic constraints in tokenization for audio understanding and demonstrate that scaling backbones fail to compensate for information loss in audio representation, especially in data-limited tasks. These findings offer practical guidance for balancing semantic density, fidelity, and efficiency in future LALMs.

cs.SD↗

arXiv2026

LoSATok: Low-Dimensional Semantic-Acoustic Tokenizer for Cross-Domain Audio Understanding and Generation

Audio tokenizers are fundamental to unifying audio understanding and generation. Understanding requires high-level semantics, while generation demands semantic and acoustic details. Existing unified tokenizers jointly encode both in high-dimensional continuous latents, which increases the modeling burden of Diffusion Transformers (DiTs) for generation. We propose LoSATok, a low-dimensional audio tokenizer for cross-domain audio understanding and generation. Motivated by the observation that 1280-dimensional semantic encoder features are compressible, we introduce a Semantic Bottleneck that compresses them into 128 dimensions, regularized by the proposed time-relation loss for temporal feature consistency. We further design a dual-level semantic supervision method that leverages both high- and low-dimensional semantic signals, enabling the tokenizer to jointly capture semantics and acoustic details within a compact latent space. Experiments on speech, music, and general audio show that SemBo preserves strong low-dimensional semantic capacity and LoSATok retains competitive understanding performance compared with several semantic representations, while consistently improving DiT modeling performance on speech, music, and audio generation. These results demonstrate that LoSATok's low-dimensional representations can effectively support audio understanding and generation. Our code is provided at https://github.com/wxzyd123/LoSATok.

eess.AS↗

arXiv2026

VISA: A Visual Information Strengthened Audio-Reasoning System for the Interspeech 2026 ARC Agent Track

Audio reasoning requires multi-step, evidence-grounded inference over temporally dynamic and acoustically mixed signals, exceeding conventional perception tasks such as ASR or captioning. We present VISA, our submission to the Interspeech 2026 Audio Reasoning Challenge (Agent Track), evaluated via the MMAR Rubrics for correctness and reasoning quality. Under a "LALM as a Tool" paradigm, VISA strengthens large audio language models with auxiliary multi-modal evidence while avoiding heavy orchestration. The system integrates three components: multi-modal feature extraction for complementary audio and acoustic-visual clues, model-voting inference with consistency checking for stable predictions, and fine-grained category-aware routing to resolve disagreements and select rubric-aligned reasoning chains. On the official Agent Track leaderboard, VISA ranks 2nd overall with a 66.23% Rubrics score. It also achieves 77.40% Accuracy, the highest among all systems listed across both the Single Model and Agent tracks.

eess.AS↗

arXiv2026

Microscopic Realization of Topologically Quantized Alignment in Fast-Rotating Nuclei

We present the first quantitative microscopic realization of topologically quantized alignment in a finite nuclear system. The realization is obtained by exact diagonalization of a cranking seniority model, with the first Chern number evaluated over the sphere of cranking-axis orientations and analyzed together with the orientation-averaged alignment and cranking-frame configuration probabilities. The Chern number changes in integer steps as the system evolves from initially paired configurations to increasingly aligned configurations. A new intermediate phase is found in which the Chern number is already nonzero while the alignment continues to evolve toward its quantized value. We show that this deviation originates from the competition among pairing, axial quadrupole splitting, and Coriolis mixing. Thus, our microscopic approach reveals a more nuanced emergence of topologically quantized alignment in realistic nuclei, providing a quantitative stepping stone toward experimental investigations.

nucl-th↗

arXiv2026

Learning Multi-Timescale Interventions under Safety and Resource Constraints

Many sequential decision problems offer qualitatively different ways of influencing the environment: some interventions act immediately, whereas others induce persistent effects that continue to shape future states long after the decision that initiated them. An agent must then decide jointly when to intervene, which temporal mode to use and how strongly, while accounting for residual effects and limited intervention resources. We introduce MINT: Multi-timescale Intervention Network Training. Persistent effects are carried by an augmented intervention state that accumulates and decays, while a structured policy separates intervention-mode selection from conditional control. Unlike temporal abstractions that extend policy execution, persistent-effect interventions remain part of the environment dynamics and may overlap with later interventions. We show that the augmented state is a sufficient statistic for the intervention history, preserving the Markov property, and establish Bellman contraction and almost-sure convergence of a tabular Q-learning instance. Across persistent-control MuJoCo locomotion, stochastic inventory management, and a physiologically grounded Type 1 Diabetes Mellitus (T1DM) simulator, MINT attains the best mean primary metric on two of three benchmarks while using fewer intervention activations than an identically augmented flat policy. Its return advantage is present at every binding intervention budget and narrows to parity in the unconstrained reference setting. In T1DM it achieves $90.9\pm0.9\%$ time in range with zero time below range, improving on the strongest baseline by $21.5\%$ points. Code, benchmarks and the configuration for every reported run are available at https://github.com/heywanrong/mint-rl.

cs.LG↗

arXiv2026

Relativistic dynamical effects in proton emission: the Wentzel-Kramers-Brillouin method for 1+1 dimensional Dirac equation

Starting from the $1+1$ dimensional (one spatial and one temporal dimension) Dirac equation, we employ the Wentzel-Kramers-Brillouin (WKB) approximation to derive the corresponding relativistic penetration probability. The derivation shows that the semiclassical momentum is determined by the Schrödinger-equivalent potential $ U_{\text{eff}}(r) = S(r) + \frac{E}{m}V(r) + \frac{S^{2}(r)-V^{2}(r)}{2m}$, instead of the simple sum of scalar and vector potentials $S(r)+V(r)$, which has been adopted widely in the studies of relativistic quantum tunneling. We then quantify the relativistic dynamical effects in proton emission by comparing the results obtained with $U_{\text{eff}}(r)$ and those obtained with $S(r)+V(r)$. Incorporating $U_{\text{eff}}(r)$ systematically reduces the penetration probability and the assault frequency, and consequently increases the predicted half-life. The relativistic dynamical effect becomes more pronounced with higher orbital angular momentum and can reach about $84\%$ in the half-life of $^{144}\mathrm{Tm}$.

nucl-th↗

arXiv2026

dots.tts.edit: Precisely Controlled Speech Editing with a Continuous Autoregressive Model

Speech editing for content creation requires precise control over both what an edit should do and where it should apply. Free-form natural language provides a flexible interface for expressing edit requests, but its ambiguity may leave the intended operation, parameters, or target region underspecified. We study a precise and explicit interface for speech editing: a transcript-grounded structural edit instruction with XML-style tags explicitly specifies typed operations and localizes them to transcript spans or boundaries. This semantic timeline avoids explicit timestamp alignment and provides an externally inspectable contract for compositional edits. We instantiate the interface in dots$.$tts$.$edit, an editor adapted from the continuous autoregressive dots$.$tts foundation model. Four representative speech-creation controls cover lexical content, affective expression, pitch and speaking-rate delivery, and temporal phrasing through text, emotion, prosody, and pause editing. Task-specific data pipelines construct operation- and scope-controlled pairs while retaining source-derived context outside each target region. We further introduce doteBench, a bilingual evaluation suite that measures precise instruction following, local preservation, and audio quality across the four controls and their composition. Experiments show leading overall instruction following and local preservation across its five editing categories, while audio quality remains comparable to existing open-source systems. Across three Seed-TTS-Eval shards, the model shows negligible differences from the base model in zero-shot TTS recognition error rate and speaker similarity.

cs.SD↗

arXiv2026

Externalizing Research Synthesis and Validation in AI Scientists through a Research Harness

AI systems can increasingly automate scientific workflows, but the reasoning that links prior evidence, generated ideas, experiments and final claims often remains implicit inside model inference. Here we introduce Xcientist, a research harness that externalizes research synthesis and experimental validation into inspectable, contract-governed processes. Xcientist organizes literature evidence, idea states, implementation plans, ablation records and repair traces as persistent research artifacts, so that generated mechanisms can be grounded, executed, tested and revised without losing their evidential basis. We identify claim drift as a failure mode of automated research, where runnable artifacts no longer support the mechanism originally claimed. Across training-free memory systems, graph-structured traffic forecasting and multi-scale physics-informed neural networks, Xcientist preserves traceable trajectories from problem formulation to mechanism design, validation and bounded revision. These results suggest that AI scientists should be evaluated not only by their final artifacts, but by whether their synthesis and validation processes remain attributable, inspectable and scientifically accountable.

cs.AI↗

arXiv2026

XFlow: An Executable Protocol Programming System for Reliable Multi-Agent Workflows

LLM-based multi-agent systems increasingly coordinate planning, reasoning, tool use, and human interaction, yet their reliability remains limited. A central source of this limitation is the underspecified prompt--harness boundary. Current systems lack a principled way to decide which workflow commitments should remain in prompts and which should become harness structure. We present \textbf{XFlow}, an executable protocol programming system for reliable multi-agent workflows, and \textbf{XPF} (XFlow Protocol Format), its domain-specific protocol programming language. XFlow occupies a middle position between prompt-only orchestration and markup-like workflow descriptions. XPF remains readable as a literate protocol, but it is compiled and executed as a program. Its design keeps informal semantic work inside actors while moving selected commitments into harness structure that can be checked, preserved, and enforced. At runtime, XFlow stages uncertainty through lifecycle-governed symbols, which are typed state cells with validation and commit states. Actor outputs are mediated before they become shared state, instead of spreading through prompts, transcripts, or implicit memory. Our experiments cover Constrained Interaction, Long-Context Reasoning, and Agentic Software Engineering. They show that XFlow improves reliability by making constraints, evidence handling, and process requirements explicit and enforceable.

cs.PL↗

arXiv2026

RAS: a Reliability Oriented Metric for Automatic Speech Recognition

Automatic speech recognition systems often produce confident yet incorrect transcriptions under noisy or ambiguous conditions, which can be misleading for both users and downstream applications. Standard evaluation based on Word Error Rate focuses solely on accuracy and fails to capture transcription reliability. We introduce an abstention-aware transcription framework that enables ASR models to explicitly abstain from uncertain segments. To evaluate reliability under abstention, we propose RAS, a reliability-oriented metric that balances transcription informativeness and error aversion, with its trade-off parameter calibrated by human preference. We then train an abstention-aware ASR model through supervised bootstrapping followed by reinforcement learning. Our experiments demonstrate substantial improvements in transcription reliability while maintaining competitive accuracy.

cs.SD↗

arXiv2026

G-STAR: End-to-End Global Speaker-Tracking Attributed Recognition

We study timestamped speaker-attributed automatic speech recognition (SA-ASR) for long-form, multi-party speech with overlap. In this setting, chunk-wise inference must preserve meeting-level speaker identity consistency while producing time-stamped, speaker-labeled transcripts. Prior Speech-LLM systems tend to prioritize either local diarization or global labeling, lacking the ability to jointly model fine-grained temporal boundaries and robust cross-chunk identity linking. We propose G-STAR, an end-to-end framework that couples a cache-conditioned speaker-tracking module with a Speech-LLM transcription backbone. The tracker provides structured speaker cues with temporal grounding, and the LLM generates attributed text conditioned on these cues. G-STAR supports component-wise optimization and joint end-to-end training, enabling flexible learning under heterogeneous supervision and domain shift. Under chunk-wise decoding protocols, experiments on both oracle-segmented local evaluation and full-meeting global evaluation show strong speaker-attributed transcription performance.

eess.AS↗

arXiv2026

A Unified and Reproducible Experimentation Framework for Speech Understanding

Speech foundation models and Speech LLMs have advanced speech understanding, yet deployment-oriented model selection is hindered by non-comparable evaluations caused by mismatched post-processing, and by training results that are hard to reproduce across data scales and pipelines. We present SURE, a unified experimentation framework that standardizes prediction formats, normalization, and scoring. SURE evaluates strong systems across paradigms, from conventional pipelines to Speech LLMs, on representative tasks under realistic acoustic and linguistic stressors. Beyond evaluation, SURE introduces an agent-assisted training conversion flow that maps paper and code into versioned, runnable training pipelines under a unified protocol on matched open-data subsets. Overall, SURE improves comparability and reproducibility for deployment-oriented evaluation.

eess.AS↗

arXiv2026

Audio-Mind: An Auditable Agentic Framework for Audio Understanding

Audio agents extend large audio-language models (LALMs) by decomposing audio questions into tool calls, intermediate evidence, and iterative reasoning steps. However, as LALMs become stronger, the key challenge shifts from enabling tool use to determining when agentic evidence acquisition genuinely benefits audio understanding. We propose Audio-Mind, an auditable and pluggable framework for conditional evidence acquisition in audio understanding. Audio-Mind dynamically combines a strong frontend with planner-guided tool use, preserving frontend judgment when initial evidence is sufficient while acquiring bounded external evidence for questions with unresolved evidence gaps. Experiments on MMAR and MSU-Bench show that Audio-Mind outperforms prior audio-agent baselines, reaching 80.4% accuracy on MMAR and 82.8% accuracy on MSU-Bench. A matched-backbone comparison highlights why this design matters: under strong audio frontends, agentic decomposition can become an orchestration bottleneck when the workflow does not preserve the frontend's holistic audio-grounded judgment. Beyond accuracy, Audio-Mind produces higher-quality, auditable reasoning traces that expose uncertainty, tool evidence, and answer rationales, offering a potential basis for more reliable audio-QA annotation and error analysis.

eess.AS↗

arXiv2026

TASU2: Controllable CTC Simulation for Alignment and Low-Resource Adaptation of Speech LLMs

Speech LLM post-training increasingly relies on efficient cross-modal alignment and robust low-resource adaptation, yet collecting large-scale audio-text pairs remains costly. Text-only alignment methods such as TASU reduce this burden by simulating CTC posteriors from transcripts, but they provide limited control over uncertainty and error rate, making curriculum design largely heuristic. We propose \textbf{TASU2}, a controllable CTC simulation framework that simulates CTC posterior distributions under a specified WER range, producing text-derived supervision that better matches the acoustic decoding interface. This enables principled post-training curricula that smoothly vary supervision difficulty without TTS. Across multiple source-to-target adaptation settings, TASU2 improves in-domain and out-of-domain recognition over TASU, and consistently outperforms strong baselines including text-only fine-tuning and TTS-based augmentation, while mitigating source-domain performance degradation.

eess.AS↗

arXiv2026

Collective quantum tunneling with time-dependent generator coordinate method

Inspired by the work of McGlynn and Simenel [Phys. Rev. C {\bf 102}, 064614 (2020)], this study investigates the quantum tunneling of two interacting distinguishable particles in two potential wells. We first benchmark the system by reproducing key established results: the exact quantum solution and the spurious self-trapping effect that arises in the real-time mean-field dynamics for strong interactions. To exactly capture the tunneling dynamics, we apply the time-dependent generator coordinate method (TDGCM) to the model. Numerical simulations demonstrate that the TDGCM, by utilizing the real-time mean-field states as generator states, successfully overcomes the self-trapping effect, yielding tunneling dynamics in excellent agreement with the exact solution. Furthermore, we explore the expectation values of the generator coordinates from the correlated TDGCM many-body wave function. While different methods for calculating expectation values show consistent results in some cases, significant discrepancies are observed in others, providing critical insights into the emergence of collective and single-particle behaviors in interacting systems. This work also verifies the TDGCM as a robust framework for describing collective quantum tunneling and opens avenues for its application to more complex and realistic systems.

nucl-th↗

arXiv2026

Beyond Descriptions: A Generative Scene2Audio Framework for Blind and Low-Vision Users to Experience Vista Landscapes

Current scene perception tools for Blind and Low Vision (BLV) individuals rely on spoken descriptions but lack engaging representations of visually pleasing distant environmental landscapes (Vista spaces). Our proposed Scene2Audio framework generates comprehensible and enjoyable nonverbal audio using generative models informed by psychoacoustics, and principles of scene audio composition. Through a user study with 11 BLV participants, we found that combining the Scene2Audio sounds with speech creates a better experience than speech alone, as the sound effects complement the speech making the scene easier to imagine. A mobile app "in-the-wild" study with 7 BLV users for more than a week further showed the potential of Scene2Audio in enhancing outdoor scene experiences. Our work bridges the gap between visual and auditory scene perception by moving beyond purely descriptive aids, addressing the aesthetic needs of BLV users.

cs.HC↗

arXiv2026

SpikeSMOKE: Spiking Neural Networks for Monocular 3D Object Detection with Cross-Scale Gated Coding

With the wide application of 3D object detection in some fields such as autonomous driving, its energy consumption is constantly increasing, making the research on low-power consumption alternatives a key research area. The spiking neural networks (SNNs), possessing low-power consumption characteristics, offer a novel solution for this research. Consequently, we apply SNNs to monocular 3D object detection and propose the SpikeSMOKE architecture, which represents a new attempt at low-power monocular 3D object detection. It's well known that the discrete signals of SNNs can lead to information loss compared to artificial neural networks (ANNs), which limits their feature representation capabilities. To solve this problem, inspired by the synaptic filtering mechanism of biological neurons, we propose a new Cross-Scale Gating Coding Mechanism (CSGC), which can enhance feature representation by combining cross-scale fusion of attentional methods and gated filtering mechanisms. In addition, to reduce the computation and accelerate training, we present a novel light-weight residual block that can maintain spiking computing paradigm and the highest possible detection performance. Our method is effective on the KITTI, NuScenes-mini and CIFAR10/100 datasets. Compared to the baseline SpikeSMOKE under the 3D Object Detection, the proposed SpikeSMOKE with CSGC can achieve 11.78 (+2.82, Easy), 10.69 (+3.2, Moderate), and 10.48 (+3.17, Hard) on the KITTI autonomous driving dataset by AP|R11 at 0.7 IoU threshold, respectively. It is worth noting that the results of SpikeSMOKE can significantly reduce energy consumption compared with the results of SMOKE. And SpikeSMOKE-L (lightweight) can further reduce the amount of parameters by 3 times and computation by 10 times compared to SMOKE.

cs.CV↗

arXiv2026

TC-BiMamba: Trans-Chunk bidirectionally within BiMamba for unified streaming and non-streaming ASR

This work investigates bidirectional Mamba (BiMamba) for unified streaming and non-streaming automatic speech recognition (ASR). Dynamic chunk size training enables a single model for offline decoding and streaming decoding with various latency settings. In contrast, existing BiMamba based streaming method is limited to fixed chunk size decoding. When dynamic chunk size training is applied, training overhead increases substantially. To tackle this issue, we propose the Trans-Chunk BiMamba (TC-BiMamba) for dynamic chunk size training. Trans-Chunk mechanism trains both bidirectional sequences in an offline style with dynamic chunk size. On the one hand, compared to traditional chunk-wise processing, TC-BiMamba simultaneously achieves 1.3 times training speedup, reduces training memory by 50%, and improves model performance since it can capture bidirectional context. On the other hand, experimental results show that TC-BiMamba outperforms U2++ and matches LC-BiMmaba with smaller model size.

eess.AS↗