Searcharxiv⌕ Search

arXiv subjects

Ziyang Zhang

Publications and source records attributed to Ziyang Zhang.

At least 19 recordsLinked to original sources

PARAGraph: Pathology-Anatomy-Aware Hierarchical Graph for Diabetic Retinopathy Grading

Diabetic retinopathy (DR) remains a leading cause of vision loss among working-age adults worldwide, making reliable severity grading clinically important. Despite strong performance, most deep models formulate DR grading as image-level classification and do not explicitly model clinically grounded evidence, such as lesion types and spatial relations. In this paper, we propose PARAGraph, a Pathology-Anatomy-Aware Hierarchical Graph framework for DR grading. PARAGraph represents each image as a three-level hierarchical graph with lesion-level nodes, intermediate category and region nodes, and global anatomical and semantic nodes. To incorporate medical priors into nodes, we construct an optic disc-fovea-anchored coordinate frame that provides a scale- and rotation-normalized retinal reference system. Within this frame, lesion nodes are encoded with category, normalized area, and anatomical coordinates. To mitigate noisy lesion segmentation, PARAGraph uses a dual-fusion strategy that introduces global visual context into a graph semantic node and a decision-level prediction branch, improving robustness when lesion evidence is unreliable. Extensive experiments on Messidor-2, APTOS, and DDR show that PARAGraph achieves consistent DR grading performance over state-of-the-art methods. Interpretability and robustness analyses further demonstrate that its predictions are clinically grounded, closely associated with lesion evidence and robust to lesion segmentation noise.

cs.CV↗

Folding Language Models into Themselves

Knowledge distillation usually trains a compact student from scratch, which needs a large amount of data even though the teacher's weights already encode most of what the student has to learn. In this work, we explore building the student directly from the teacher's weights. We propose Concertina, which folds every group of N adjacent teacher layers into one shared block that is executed N times. Each loop step carries a low-rank correction initialized from the layer it replaces, and a per-token gate mixes these corrections, to let every token recover the behavior of the original layers from a single shared block. We further design a distillation data recipe that balances general, mathematical, academic and targeted multiple-choice data, to recover all capabilities of the teacher together rather than one at the expense of another. On Qwen3-4B, a 2.27B Concertina model distilled on 22B tokens retains about 90% of the teacher on commonsense and mathematical reasoning and 84% on average across eight benchmarks. It outperforms same-size dense and layer-sharing baselines, as well as Qwen3-1.7B, which is trained on over a thousand times more data. These results suggest that a compact model need not be retrained at scale, since folding the teacher into itself and distilling it on the right data recovers most of its capabilities. Code is available at https://github.com/myboker/concertina.

cs.CL↗

Report of the 8th LSVOS Challenge: Mobile Unified Multimodal Understanding

This report summarizes the Mobile Unified Multimodal Understanding (MUMU) track of the 8th Large-scale Video Object Segmentation (LSVOS) Challenge, held in conjunction with ECCV 2026. MUMU asks for a single unified model that jointly performs multi-concept image tagging over quality, scene, and event vocabularies, open-vocabulary object detection, and concise image captioning, within a budget of 0.5B parameters and 8 GB of peak inference memory. Submissions are ranked by a weighted task score scaled by an efficiency factor. We describe the tasks, rules, and evaluation protocol, report the final results, and review the methods of the top three teams. All three leading solutions build on the compact Florence-2-base model, but extend it in different ways: lightweight tagging heads with cross-task inference refinement, a shared low-rank adapter trained on coverage-aware multi-source data with task-dependent decoding, and a training-free pipeline that maps captions to tags and applies detection test-time augmentation. The results show that a compact prompt-based vision-language model can serve all three outputs when supervision, decoding, and cross-task cues are designed for each task, while open-vocabulary detection on novel categories remains the main bottleneck.

cs.CV↗

Discovering Effective Agent Memory Designs through Program Evolution

Program evolution provides a promising approach to discover agent memory designs and study the choices underlying their effectiveness. We introduce MAPE, a framework that evolves whole memory programs from minimal seeds through a standardized lifecycle interface and an iterative diagnosis-proposal loop. Experiments on LoCoMo and LongMemEval yield an absolute accuracy gain of up to 12% over manually designed baselines at 35% of the serving cost, with further evaluations supporting transfer across serving backbones and benchmarks. Empirical analyses and ablations identify and validate two recurring design patterns: retaining raw conversations as retrievable evidence preserves details omitted from extracted memories, while composing retrieved evidence according to query intent helps the answering model use it effectively. Transplanting the composition module into five existing memory systems improves accuracy across all five. Beyond these patterns, the search objective shapes where computation goes: optimizing serving cost alongside accuracy shifts computation from LLM-based memory construction to selective processing at query time while retaining competitive accuracy. Together, these findings offer empirical lessons for agent memory design. Code is available at https://github.com/ash-ding/MAPE.

cs.CL↗

What Helps Denoising Can Hurt Memory: Rethinking History in Long Video Generation

Autoregressive video generators produce long videos block by block, using historical context to maintain temporal consistency. After denoising, a committing pass encodes each clean block into keys and values (KV) that later blocks read, and the same clean block can be encoded into different KV depending on the history it reads. We therefore study historical context not only as a condition for denoising but also as an input to committing. Older historical context that supports denoising can harm subsequent generation when read during committing. At fixed context size and temporal positions, older historical context lowers mean clip-to-clip Subject Consistency. Across 5 generator configurations supporting longer contexts, consistency rises, plateaus, and then declines as committing context expands. Motivated by these findings, we propose Commit Forcing, a training-free rule that organizes committing into a chain. During committing, each block reads only the preceding block's stored KV, so older historical context reaches new entries only through the chain. We apply this rule to 6 pretrained generators. On 60-second VBench, Commit Forcing improves Quality and Total on all 6 generators, by up to 1.27 Quality points and 1.10 Total points. Long-range consistency gains extend to 240-second videos. In a blinded comparison on 120-second LongLive videos, Commit Forcing achieves win rates of 83.5% for subject and scene consistency and 82.9% for overall preference against the base generator, excluding ties.

cs.CV↗

GeoAgent: Contract-Constrained LLM Agents for Seismic Data Processing

Large language models (LLMs) can drive scientific software through function calling, but free-form tool use is fragile on long, data-dependent processing chains. Models omit arguments, invent units, feed the previous artifact into the wrong input, and report completion without a verifiable product. We present GeoAgent, an agent system for seismic data processing built on Madagascar's roughly 1{,}965 command-line programs. GeoAgent exposes 54 operator agents (51 generated from a declarative specification), retrieves from an index of 1{,}937 operators, and encodes the processing flow as a nine-stage state graph. Four explicit contracts---named artifacts, input resolution, parameter takeover, and artifact validation---move everything the graph can determine out of the model's discretion. A deterministic geometry tool parses the self-describing SPS/RPS/XPS and SEG-Y headers, cross-checks eight consistency conditions, and reproduces the reference shell recipe exactly. The system runs in containers with session- and task-level isolation, bounded execution, and defenses against silent failures. On a 2-D land dataset, the nine-stage pipeline completes in about six to eleven minutes and yields a verified artifact at every stage, while retrieval reduces per-turn tool context from about 12k to 2k tokens. We report the design, deployment experience, and motivating failure modes.

physics.geo-ph↗

MMTA: Benchmarking Multimodal Temporal Analysis with Time Series, Text, and Vision

Time-series analysis is increasingly studied in multimodal settings, yet most benchmarks pair temporal signals with text or with visualizations derived from the same signal. This leaves open whether multimodal large language models (MLLMs) can integrate time series with real-world visual evidence. We introduce MMTA (multimodal temporal analysis), a benchmark of tri-modal tasks with diverse cross-modal complementarity patterns among time series, text, and vision. MMTA spans nine domains -- robotics, autonomous driving, human action, manufacturing, AR assistance, healthcare, recommendation, finance, and climate -- under video+TS and image+TS layouts, with a 3,524-sample test split covering classification and forecasting. The benchmark standardizes heterogeneous real-world sources into a shared schema with domain-specific prompts, provenance documentation, and success-weighted metrics that penalize both wrong answers and failed generations. Evaluating eight open-source and proprietary MLLMs reveals a shared bottleneck: representing time series as text creates very long token sequences and weakens fine-grained alignment with visual evidence, especially in long-video settings. We then introduce TimeOmni-v, a multimodal LLM extension with a dedicated time-series encoder, temporal positional alignment, modality-aware token layout, and a forecasting head. TimeOmni-v improves long-video classification and time-series inference efficiency, showing that MMTA not only exposes representation and grounding failures but also motivates targeted model design.

cs.LG↗

TVPara: Paralinguistic Captioning of Expressive TV-Series Speech in Complex Acoustic Environments

Large audio-language models (LALMs) struggle to parse paralinguistics (e.g., emotion, pace, pitch) in speech that combines intense emotion with complex acoustic environments. Strengthening this ability calls for training data that reflects such conditions. However, existing corpora rely heavily on studio-recorded speech, offering limited emotional and acoustic diversity. To address this gap, we introduce TVPara, a paralinguistic captioning dataset built from TV series. Speech in TV series often carries diverse emotional expression and co-occurs with scene-specific background sound, providing expressive emotion and complex acoustic environments at scale. Constructed via LALM pre-annotation and human correction, TVPara includes two tasks: Single-Utterance (~492h) and Multi-Utterance (~213h) Paralinguistic Captioning. Fine-tuning open-source LALMs on TVPara yields an average relative improvement of 31.26% in paralinguistic understanding under complex acoustic environments. The gains further generalize to an unseen TV series and to cleaner, cross-domain speech.

cs.SD↗

Time-JEPA: Grounding Video World Models with Physical Time Series

Video world models learn how visual environments evolve by predicting missing or future content. However, physical-state information can provide useful context for predicting how environments evolve. Prior work incorporates this information through different mechanisms: domain-specific physical laws can complicate multi-domain training, while supervision on individual frames or frame pairs gives limited attention to how physical state evolves over time. We introduce Time-JEPA, which fine-tunes a video world model using low-dimensional, temporally aligned physical time series as auxiliary supervision. Physical time series connect video representations with physical state at corresponding temporal positions, encouraging the model to preserve how physical state evolves. In multiple domains including autonomous driving, robot manipulation, and human motion, Time-JEPA improves current-state decoding, multi-horizon future-state prediction, and kinematic reconstruction over matched video-fine-tuned controls across both condition and state-variable distribution shifts. Multi-domain training provides aggregate gains over matched single-domain training in all three tasks. External evaluations further suggest that time-series supervision improves sensitivity to physical anomalies and the representation of physical interactions, including in unseen scenarios. Aggregate gains on a second video backbone indicate that the method can be applied to more backbones. These results suggest that physical time series help video world models better understand the physical world.

cs.CV↗

TimeOmni-o1: Bidirectionally Coupling Time Series Foundation Models and LLMs for Multimodal Reasoning and Forecasting

Time series from real-world systems are often accompanied by text, such as a report on an event that affects them. Two common tasks on these series are understanding, e.g., answering questions about a series, and forecasting its future values. Both tasks require reasoning over the series and the text together. Moreover, understanding outputs text while forecasting outputs a time series, so serving both requires a model that reads and generates both modalities. Large language models (LLMs) read time series as digit tokens and handle long series poorly. Time series foundation models (TSFMs) forecast well from large-scale pretraining but typically neither read nor generate text. Time series language models add a time series modality to the LLM but forecast on the LLM side, leaving the TSFM's pretrained forecasting ability unused. We present TimeOmni-o1, which couples a pretrained TSFM and an LLM bidirectionally. In one direction, the TSFM encodes the series in compressed form for the LLM, which reasons over it together with the text. In the other, the LLM's reasoning is fed back into the TSFM, which produces the forecast conditioned on it. One model architecture thus reads and generates both series and text, reasons in text, and draws on the TSFM's pretrained forecasting ability. To train it, we build MMTR, a multimodal time series reasoning dataset with annotated reasoning traces, covering both tasks across 12 domains with 500k+ samples. On MMTR, TimeOmni-o1 outperforms existing models on both understanding and forecasting across most domains, as well as the LLM and the TSFM inside it when used individually. The gains carry over to three open benchmarks, where it outperforms the dedicated time series models.

cs.CL↗

Bayesian Speech Synthesizers Can Learn from Multiple Teachers

Text-to-speech (TTS) is a "one-to-many" problem: one text can be spoken in many valid ways, so its realizations form a distribution whose spread depends on the input. Most TTS models do not model this spread explicitly: they leave it implicit in the sampling process or fix it in advance. To model this distribution explicitly, we propose BELLE (Bayesian evidential learning with language modelling), which places a Bayesian prior over the mean and variance of each mel-spectrogram frame and predicts this prior from the input, while keeping the architecture and single-pass sampling of common continuous-valued autoregressive TTS models such as MELLE. Estimating this variance requires several recordings of each text, whereas standard TTS corpora provide one. We therefore construct two parallel-speech corpora: ParallelVox from real human recordings and LS+6TTS, which adds speech from six pre-trained TTS teachers to LibriSpeech (LS). Trained on these corpora, BELLE outperforms MELLE under the same training conditions, is competitive with larger TTS systems on the objective metrics of LibriSpeech test-clean, and supports streaming generation. Analyses of repeated human readings show that real speech has the properties the Bayesian prior assumes and that the variance predicted by BELLE follows the variation across these readings. Parallel speech improves BELLE, and synthetic parallel speech performs comparably to real parallel speech at the same volume and is far easier to scale.

cs.SD↗

FreeLoc: Online Floorplan Localization via Diffusion-Aided Pose Refinement

Floorplans provide compact and widely available geometric maps for indoor localization, but existing high-performing floorplan-based methods still convert them into dense scene-specific offline databases, tying accuracy, storage, and runtime to the sampling resolution of the discretized pose space. We present FreeLoc, an online RGB-based floorplan localization framework that treats the floorplan as a directly queryable geometric map. FreeLoc introduces an efficient online geometric querying and diffusion-aided refinement scheme, which retrieves plausible pose anchors through on-the-fly floorplan ray querying and refines them into accurate continuous pose estimates. For sequential localization, FreeLoc develops an online likelihood construction strategy that bridges single-frame localization and probabilistic temporal fusion by constructing likelihoods from coarse-sampled candidates and refined pose hypotheses, enabling histogram-filter-based temporal fusion without offline databases. Experiments demonstrate real-time online inference and state-of-the-art performance in both single-frame and sequential localization, while real-world results validate practical deployability in indoor robotic localization scenarios.

cs.RO↗

Cross-Modal Solar Image Synthesis: Adapting the Surya Foundation Model from He I 10830 Å to EUV Translation and Coronal Hole Segmentation

The long observational record of He I 10830 Å offers a means to investigate solar morphology before modern extreme-ultraviolet (EUV) imaging. We adapt the Surya solar foundation model to predict Solar Dynamics Observatory/Atmospheric Imaging Assembly (SDO/AIA) 94, 193, and 304 Å images and a coronal hole (CH) probability map from full-disk helium observations. A convolutional input adapter, low-rank backbone updates, and dedicated output decoders learn from temporally paired, geometrically registered observations, with Spatial Possibilistic Clustering Algorithm (SPOCA) catalog polygons providing CH supervision. On observations held out from downstream fine-tuning, the selected dedicated models achieve disk-restricted correlation coefficients (CCs) of 0.8196, 0.8885, and 0.8398 for the three AIA channels, respectively; the CH model achieves an intersection over union (IoU) of 0.4360. The predictions recover broad solar structure, although local agreement varies substantially by channel. An optional residual refiner addresses spatial detail, and its application on pre-SDO dates improves the correlation of synthetic AIA 304 with Solar and Heliospheric Observatory/Extreme-ultraviolet Imaging Telescope (SOHO/EIT) 304 references from 0.6511 to 0.7085. Together, these results support the feasibility of helium-conditioned EUV morphological proxies and motivate their use in historical reconstruction, within the scope of the downstream test and cross-instrument evaluation.

astro-ph.SR↗

XMatch: Enhancing Covariate-Aware Time Series Forecasting through Tree-Structured Exogenous Matching

Future exogenous variables provide valuable information for forecasting endogenous time series. Existing covariate-aware methods primarily learn the direct influence of exogenous variables on endogenous variables. However, these effects can be complex and change with the pattern of the exogenous variables, making them difficult to capture. Beyond this perspective, we observe that a given exogenous pattern often co-occurs with only a small set of endogenous response patterns. These associations motivate a strategy that matches future and historical exogenous patterns and uses the corresponding endogenous patterns to enhance forecasting. However, in real-world forecasting scenarios with multiple exogenous variables, each exogenous variable provides a distinct dimension for matching, creating a dilemma for this strategy between precise matching and sufficient historical support. To bridge this gap, we propose XMatch (EXogenous MATCHing), a covariate-aware forecasting model that realizes the aforementioned strategy through a tree-structured matching process that adaptively adjusts the number of exogenous variables used as matching conditions. Specifically, we first introduce the ProtoTree Creator, which organizes historical correspondences between exogenous and endogenous patterns into a ProtoTree, whose deeper levels incorporate additional exogenous variables for matching. For forecasting, we then design the ProtoTree Matcher, which uses future exogenous variables to query the ProtoTree and adaptively determines how many exogenous variables to use for matching based on exogenous pattern similarity and historical support. Finally, the matched endogenous patterns are used as explicit historical evidence to enhance forecasting. Extensive experiments on 12 real-world datasets demonstrate that XMatch outperforms state-of-the-art baselines.

cs.LG↗

Post-Training Leaves Behavioral Shadows on Unrelated Decisions

We find that language models can transfer capabilities through task-unrelated text. Post-training typically improves language models using task-specific data. Prior work on subliminal learning shows that information about these updates can pass through unrelated generations, but has largely focused on traits or preferences using extensive teacher outputs. We introduce Active Taskless Distillation (ATD), which achieves capability transfer using only a single word from the teacher per prompt. ATD probes the behavioral shadow of post-training by selecting prompts where the teacher and student's shared public ancestor is nearly indifferent between two ordinary words. A student initialized from this ancestor learns solely from the resulting prompt-word pairs, without target-task examples, teacher logits, or teacher parameters. In the primary coding experiment with Qwen2.5-1.5B, 5,664nses yield a 5.34 pp gain on HumanEval+ over an exact nuisance-matched control thadisrupts prompt-resperiments showtransfer in scientific knowledge, commonsense reasoning, and reading comprehensins across additional model generations, sizes, and families. Functional analyses show that the learned sid composable, andthat its strength tracks the teacher's update strength.

cs.CL↗

Physics-Informed Neural Networks for Fast Multilayer Spectral Inversion of Hα 6562.8 A and Ca II 8542.1 A Spectra

Strong chromospheric absorption lines such as H$α$ 6562.8 A and Ca II 8542.1 A provide vital diagnostics of plasma dynamics and thermal structure in the solar chromosphere. Multilayer spectral inversion (MLSI) offers a physically interpretable framework for modeling these lines using a finite number of radiative-transfer layers, but conventional MLSI relies on pixel-by-pixel nonlinear least-squares fitting, making it computationally expensive for large imaging spectroscopic data sets. Here, we introduce a physics-informed neural-network (PINN) framework to accelerate MLSI while preserving its analytic radiative-transfer formulation. The network predicts MLSI parameters directly from observed line profiles and passes them through a differentiable MLSI forward model to synthesize spectra. Training follows a two-stage approach: an initial stage optimized solely via spectral reconstruction loss, followed by fine-tuning that combines spectral consistency with parameter-space supervision from conventional MLSI results on a single reference image. This strategy eliminates the need for large precomputed training sets while maintaining physical interpretability. Applied to Fast Imaging Solar Spectrograph (FISS) observations from the Goode Solar Telescope (GST) targeting both quiet-Sun and active-region regions, MLSI-PINN parameter maps reproduce the primary spatial structures of direct inversions, achieving an arithmetic mean pixel-wise Pearson correlation coefficient of 0.933 across all evaluated parameters. The reconstructed spectra closely match both observed profiles and conventional MLSI fits. Post-training, MLSI-PINN processes a raster in approximately 5-15 seconds compared to 3-5 minutes for conventional MLSI, delivering an inference speedup of about 12-60 times without substantial loss in reconstruction quality, enabling efficient MLSI analysis on large chromospheric data sets.

astro-ph.SR↗

Enhancing Event Candidate Acquisition for Event Linking

Event linking associates event mentions in text with entries in a knowledge base (KB), or identifies them as out-of-KB events. Although existing methods use different architectures, candidate event acquisition can still be weakened by short ambiguous mentions, noisy arguments, and evidence that is unevenly useful for retrieval. We present MACE, a Multi-Agent Candidate Event acquisition method that refines event structure before linking. MACE uses evidence-specialized LLM agents to acquire time, location, participant, and event-type evidence, exposes intermediate queries to candidate-event lookup tools, and lets a coordinator revise the evidence set before final candidate construction. Experiments on two event linking benchmarks show that adding MACE to different event linking models consistently improves accuracy. These results show that MACE improves event linking through better candidate event acquisition without modifying the event linking model.

cs.AI↗

EpiQAL: Benchmarking Large Language Models in Epidemiological Question Answering and Reasoning

Reliable epidemiological reasoning requires synthesizing study evidence to infer disease burden, transmission dynamics, and intervention effects at the population level. Existing medical question answering benchmarks primarily emphasize clinical knowledge or patient-level reasoning, yet few systematically evaluate evidence-grounded epidemiological inference. We present EpiQAL, to our knowledge the first diagnostic benchmark for epidemiological question answering over research literature, comprising three subsets built from open-access articles across diverse diseases. The three subsets progressively test factual recall, multi-step inference, and conclusion reconstruction under incomplete information, and are constructed through a quality-controlled pipeline combining taxonomy guidance, multi-model verification, and difficulty screening. Experiments on fifteen models spanning open-source and proprietary systems reveal that current LLMs show limited performance on epidemiological reasoning, with multi-step inference posing the greatest challenge. Model rankings shift across subsets, and scale alone does not predict success. Chain-of-Thought prompting benefits multi-step inference but yields mixed results elsewhere. EpiQAL provides fine-grained diagnostic signals for evidence-grounding, inferential reasoning, and conclusion reconstruction.

cs.CL↗