Search arXiv⌕ Search

arXiv subjects

Xiaomeng Li

Publications and source records attributed to Xiaomeng Li.

At least 19 recordsLinked to original sources

Verifier Noise Is a Temperature: Testing Tilt Predictions in Finite-Budget LLM-Guided Search

LLM-guided search loops propose a modified candidate, score it, and keep it if it scores better. When both candidates are re-scored at every comparison, as in the minibatch acceptance gates of prompt optimizers, keep-if-better is a reversible Markov chain if the score difference is logistic and the mutation operator is reversible, and its stationary law is the mutation equilibrium tilted by exp(beta f). For any continuous symmetric noise difference the same tilt is the stationary law of a small-step diffusion limit, with beta = 4F'(0); for a minibatch gate beta grows with the square root of the minibatch size. Caching the incumbent's score instead inflates the recorded score. We test the tilt as a prediction for LLM-guided program search, committing each prediction from calibration data before the search runs. The level after 150 steps rises with the minibatch size in six of seven settings, but its predicted value fails: on bin packing, search attains 62-88% of the predicted gain when the proposer edits the parent and 109-119% when it writes new heuristics. The attained fraction falls with the parent-child quality correlation of unselected mutation; a registered test on two new settings confirmed the direction, but not the quantitative agreement after a correction. Further registered tests found that runs started from a good program kept losing quality over 300 steps, more so from starts that search had selected, by a margin that passed its registered rule narrowly. Runs from different starts had not reached a common level, so whether the tilt describes the stationary law of LLM editing remains open. A registered test in prompt optimisation on frozen task-model scores failed: search exceeded the tilt, and the bin-packing relation did not transfer.

cs.LG↗

Concept-Grounded Reasoning with Prompt-Driven Localization for Interpretable Structured Report Generation

Medical imaging modalities such as ultrasound and X-ray are widely used in clinical practice, where diagnosis follows a structured, evidence-driven workflow aligned with standardized criteria. While multimodal large language models (MLLMs) show promise for automated medical report generation, most existing systems rely on end-to-end multimodal fusion without modeling clinically defined intermediate attributes, leading to limited grounding and interpretability. To address this issue, we propose CORAL (COncept-grounded ReAsoning with Localization), a multimodal framework that integrates spatial grounding and concept-level supervision into a unified reasoning process. CORAL employs a prompt-driven medical segmentation model to localize lesions and predicts multi-class clinical attributes through a Concept Bottleneck module. The resulting textual concept tokens are combined with mask-modulated visual features within an MLLM to enable structured report generation and diagnostic prediction. Experiments on BUS-CoT and IU X-ray datasets demonstrate consistent improvements in diagnostic accuracy, concept consistency, and report quality over strong general-purpose and medical MLLMs, indicating that concept-grounded reasoning better aligns generation with clinical decision processes.

cs.CV↗

Multimodal Generative AI Assists Comprehensive Echocardiographic Interpretation

Echocardiography is the cornerstone of cardiovascular diagnostics, yet its global utility is constrained by a critical shortage of experienced interpreters. Current artificial intelligence (AI) solutions remain limited, typically operating as opaque "black boxes" or focusing on narrow tasks that fail to capture the holistic, guideline-driven reasoning of a clinician. Here we introduce EchoChat, a generative multimodal vision-language model that performs comprehensive echocardiographic analysis to support clinician interpretation. By emulating the cognitive workflow of a cardiologist, EchoChat integrates vast medical knowledge with a chain-of-thought reasoning mechanism to explicitly ground its conclusions in visual evidence. Trained on echocardiography-instruction pairs from 46,358 examinations, our model establishes a new state-of-the-art across diverse cohorts from six international medical centers. It supports the broadest diagnostic scope of 33 cardiac conditions and findings, including 15 newly added conditions for which it achieves an average F1 of 73.7% across available external evaluations. It surpasses existing best-performing foundation models in cardiac condition recognition (improving average recall by 13.54% on shared tasks) and quantification (average R2 score increase of 0.275 on shared tasks). Furthermore, in a source-blinded echocardiographer reader study, echocardiography reports co-generated with EchoChat were rated higher in diagnostic accuracy and completeness by two senior echocardiographers, and improved diagnostic performance from 72.78% to 82.75% in terms of F1 score over routine-reader only reports. These results demonstrate that integrating clinical reasoning into multimodal AI can democratize specialist expertise and transform the standard of cardiovascular care.

cs.CV↗

Small Average Error, Wide Individual Error: General-Purpose LLMs Estimating Energy Intake from Meal Photos in a Feeding Trial

Photo-based estimation of food energy could ease dietary assessment, and general-purpose multimodal large language models (LLMs) make such estimates easy to obtain. Most evaluations use one photo taken before eating and compare with energy served, while dietary assessment needs energy eaten. We evaluated eight models from three vendors (Claude Haiku, Sonnet and Opus 5.5; GPT-6 luna, 6.1 sol and 6 astra; Gemini 3.8 Flash and 4 Argon) on 802 meals from ACETADA, a controlled feeding trial with weighed intake, with the before photo only, before and after photos in separate calls (subtraction), or both photos in one request. Three results ran against intuition. Subtraction under-estimated intake for six of eight models, because the models still report hundreds of kilocalories for a nearly empty tray. Scaling the before-only estimate by the average share eaten (0.87) matched much of the gain from the after photo; beyond that constant, the second photo lowered meal-level error by only 2.2 to 4.4 points, mainly in meals with food left. With one photo, two models had a near-zero intake bias only because they under-estimated the served energy and ignored uneaten food. In ablations, adding the after photo improved accuracy under both prompt formats for seven models, while prompt effects varied across models. Over 148 feeding days, five models kept the 90% confidence interval of their mean error within a pre-set 10% margin; one model also met the pre-set 30% limits of agreement, narrowly (-28% to +27%; limits outside 30% cannot be excluded), and the other seven did not. Protein was under-estimated by every model, most in protein-dense meals; repeated calls changed error by at most 1.4 points; a high reasoning effort helped only the smallest model. A low average bias is not evidence of accurate individual intake estimation; such estimates should support, not replace, assessment by a dietitian.

cs.HC↗

PRET is a few-shot system for pan-cancer recognition without example training

Pathological examination stands as the cornerstone in cancer diagnosis, impacting millions worldwide annually. With the shortage of pathologists globally, artificial intelligence (AI) has emerged rapidly to automate the diagnostics process. However, conventional AI models require substantial labeled data for each disease, posing huge challenges in scalability and practicality. Therefore, we introduce PRET (pan-cancer recognition without examples training), a few-shot system to achieve flexible, scalable, and effective cancer recognition across diverse organs, hospitals and tasks without training. Evaluated on 23 international benchmarks comprising 4,484 whole-slide images, our method outperforms existing approaches across 20 tasks, achieving over 97% area under the curve on 15 benchmarks with a maximum improvement of 36.76%. Notably, PRET delivers clinical-grade diagnostic performance in lymph node metastasis detection using only eight slide examples, outperforming 11 pathologists. By offering a flexible and cost-effective solution for pan-cancer recognition, PRET paves the way for accessible and equitable AI-based pathology systems, particularly benefiting minority populations and underserved regions.

cs.CV↗

Synergistic Vision-Language Reinforcement Enables Scalable On-Demand Analysis across Diverse Clinical Tasks

Accurate segmentation of tumors and organs-at-risk is essential for oncology but remains time-consuming and expertise-intensive. Existing AI systems often require manual spatial prompts or task-specific retraining, while generic class labels provide limited guidance for heterogeneous disease targets. Here we present SyRe, a text-prompted segmentation foundation model based on Synergistic vision-language Reinforcement, which strengthens bidirectional interaction between visual and linguistic representations. We introduce Color Region Description (CRD) to generate region-level semantic supervision and construct SyReData, comprising 20 million image-mask-description triplets across 9 modalities and 229 segmentation tasks. Training with diversified prompt formats enables single-target, multi-target and open-ended segmentation, together with invalid-prompt rejection. SyRe achieves accurate segmentation across diverse clinical scenarios, with strong gains on disease-related targets. Evaluation on 28 unseen external datasets, including 17 cancer types, 3 non-cancer diseases and multinational in-house cohorts, demonstrates generalization under distribution shift. Beyond spatial overlap, SyRe-generated masks preserve quantitative pathology information. Across five retrospective tumor imaging cohorts spanning CT and MRI, radiomics derived from these masks stratify survival and improve prognostic modeling over clinical baselines. Clinician-in-the-loop refinement further supports case-level correction. These results establish SyRe as a generalizable framework for scalable quantitative oncology analysis and clinician-guided segmentation.

cs.CV↗

Neurons++: A Unified Visual-Stream Framework for Image and Video Reconstruction from fMRI with Neural-to-Visual Interpretability

Decoding visual experience from fMRI is important for understanding human perception and developing brain-computer interfaces. Recent generative decoding methods can reconstruct plausible images and videos, but visual plausibility alone does not ensure that the output is faithfully grounded in brain activity. In this work, we propose Neurons++, a visual-stream fMRI reconstruction framework for both static image and dynamic video reconstruction. Compared with our conference version Neurons, Neurons++ extends the paradigm from video-only reconstruction to visual-stream reconstruction for both images and videos, and replaces parallel task conditioning with serial cortex-inspired decoding. The model first uses image-fMRI pretraining to obtain a stable spatial-semantic initialization, and then performs subject-specific fine-tuning for image and video reconstruction separately. The serial design organizes fMRI-derived representations into explicit spatial, semantic, language-level, and temporal cues. Extensive experiments show that Neurons++ achieves state-of-the-art video reconstruction and competitive image reconstruction across standard benchmarks. Beyond reconstruction fidelity, neural-to-visual analyses indicate that Neurons++ exhibits qualitative directional controllability under neural perturbation, that its fMRI encoder read-out is sensitive to stimulus-relevant cortical features, and that it supports ROI-guided semantic modulation. Code and model weights are available at: https://github.com/xmed-lab/NeuronsPlusPlus.

cs.CV↗

MedHorizon: Towards Long-context Medical Video Understanding in the Wild

Medical multimodal large language models (MLLMs) have advanced image understanding and short-video analysis, but real clinical review often requires full-procedure video understanding. Unlike general long videos, medical procedures contain highly redundant anatomical views, while decisive evidence is temporally sparse, spatially subtle, and context dependent. Existing benchmarks often assume this evidence has already been localized through images, short clips, or pre-segmented videos, leaving the retrieval-before-reasoning problem under-tested. We introduce MedHorizon, an in-the-wild benchmark for long-context medical video understanding. MedHorizon preserves 759 hours of full-length clinical procedures and provides 1,253 evidence-grounded multiple-choice questionsthat jointly evaluate sparse evidence understanding and multi-hop clinical reasoning. Its evidence is extremely sparse, with only 0.166% evidence frames on average, requiring models to search noisy procedural streams before interpreting and aggregating findings. We evaluate representative general-domain, medical-domain, and long-video MLLMs. The best model reaches only 41.1% accuracy, showing that current systems remain far from robust full-procedure understanding. Further analysis yields four key findings: performance does not scale reliably with more frames, evidence retrieval and clinical interpretation remain primary bottlenecks; these bottlenecks are rooted in weak procedural reasoning and attention drift under redundancy, and generic sampling methods only partially balances local detail with global coverage. MedHorizon provides a rigorous testbed for MLLMs that retrieve sparse evidence and reason over complete clinical workflows.

cs.CV↗

ULF-Synth: Physics-Guided Ultra-Low-Field MRI Enhancement for Pediatric Neuroimaging

Ultra-low-field (ULF) MRI offers portable and accessible neuroimaging but suffers from reduced signal-to-noise ratio and limited spatial resolution compared to high-field (HF) systems. Acquiring paired ULF-HF data for supervised enhancement is often difficult, particularly in resource-limited settings. We introduce ULF-Synth, a framework that combines: (i) acquisition-based synthesis of realistic ULF images from HF volumes to create large-scale paired training data, (ii) a spatial-frequency domain objective that prioritizes recovery of high-frequency anatomical detail. This formulation is architecture-agnostic, consistently improving structural similarity and perceptual fidelity across encoder-decoder, adversarial, and diffusion-based translation models. When trained exclusively on synthetic data, the resulting models generalize effectively to real 64mT ULF acquisitions, improving downstream multiclass brain segmentation and achieving higher radiologist preference and diagnostic acceptability in a blinded reader study. These findings demonstrate that synthetic paired supervision provides a practical and scalable pathway for enhancing ULF MRI without requiring real paired acquisitions. Code, Models and Dataset: https://github.com/toufiqmusah/ULF-Synth

cs.CV↗

Non-Polynomial Wave Computation through Recurrent Resonant Scattering

Structural nonlinearity enables nonlinear input--output mappings to emerge from otherwise linear wave dynamics. Repeated interactions with an input-encoded structure can enhance such mappings, but finite-depth implementations restrict the accessible functional order. Here we show that recurrent scattering in a resonant cavity provides a distinct regime for structural nonlinearity. Repeated interactions are coherently accumulated into a non-polynomial response to structural perturbations. The resulting mapping naturally takes a Kolmogorov--Arnold form: perturbation-induced resonance shifts realize the inner univariate mappings, while the resonant spectral response provides the outer mapping. We identify two complementary physical controls of representation capacity: the resonance linewidth governs the functional richness within each branch, whereas combining multiple branches expands the accessible function space. Microwave-cavity measurements validate the two-stage mapping and demonstrate a two-branch nonlinear computation through XOR classification. Our results connect recurrent resonant scattering with controllable non-polynomial computation in linear wave systems.

physics.optics↗

SCIRIGOR:Evaluating Open-Ended Scientific Analysis Beyond Final Scores

Scientific coding agents produce interdependent code, results, figures, and claims, yet evaluating final outputs alone does not establish whether their conclusions are scientifically supported. We formulate evidence-grounded multimodal scientific analysis, requiring agents to produce executable analyses and claims supported by results and visualizations from the same run. We introduce SciRIGOR, an evaluation framework and benchmark comprising 100 cases from scientific articles across six domains and 17 subfields. The framework reconstructs typed evidence graphs, separates artifact fidelity from relational validity, and scores complete claim-support paths while localizing the earliest unsupported relation. Source-grounded alternative paths accommodate scientifically equivalent analyses and visualizations. We evaluate 11 agent/model configurations. On full-benchmark runs, claims agree with faithful and unfaithful results at nearly identical rates (91.8% versus 91.0%). Yet no system exceeds 62.6% on the soft evidence-chain score or 18.0% strict whole-chain success. These findings show that internal coherence does not establish scientific correctness: evaluation must verify support along the complete data-to-claim path.

cs.AI↗

PathoArgus: Advancing Evidence-Grounded Long-Context Visual Reasoning across Gigapixel Whole-Slide and Multi-Slide Case Contexts

Whole-slide pathology reasoning requires models to integrate gigapixel-scale visual evidence across complete case-linked slides, yet current question-answering benchmarks primarily measure final answer accuracy--a metric vulnerable to linguistic priors and benchmark regularities, and insufficient to establish that predictions are grounded in the supplied tissue. We introduce PathoArgus-Bench, a benchmark and evaluation protocol that explicitly tests the full evidence chain: availability, accessibility, use, and responsiveness. PathoArgus-Bench comprises 22,078 four-choice questions from 4,913 patients across 15 TCGA projects, covering six pathology capabilities across three levels of evidence demand, and operates under a fixed reader budget that retains only a small fraction of the gigapixel context. To further isolate evidence-grounded reasoning, we contribute ESG (Evidence State Quartets), a controlled set of 483 quartets where the question text is fixed while the target WSI set is moved, replaced, or removed, requiring consistent predictions across all states. Evaluating 20 general-purpose, medical, and pathology-specific systems reveals a stark gap: while GPT-5.6 achieves 57.09% overall accuracy and 57.04% on ESG, it correctly completes only 19 of 483 quartets (3.93% QExact), exposing that row-level accuracy does not translate into reliable evidence grounding. We also introduce PathoArgus, a fixed-budget reader that allocates context via question relevance and spatial coverage, attaining 50.39% overall accuracy yet only 1.86% QExact--demonstrating that improved context access alone does not ensure consistent evidence-based prediction. Our benchmark and diagnostics establish that acquiring useful whole-slide context is necessary but far from sufficient, and call for a shift from answer-centric to evidence-grounded evaluation in computational pathology.

cs.CV↗

CACSurv: Concordance-Aligned Comparative Learning with Large Language Models for Cancer Survival Prediction

Cancer survival prediction supports treatment planning, risk stratification, and follow-up management. Existing methods use structured clinical variables, whole-slide images, genomic profiles, or multimodal inputs, while patient reports remain underexplored. We study report-centric survival prediction using reports that organize pathological, clinical, and molecular evidence. Large language models (LLMs) can reason over such reports, but case-wise time regression introduces two mismatches. First, a formulation mismatch arises because survival evaluation depends on ordering comparable patients, whereas independent time predictions do not enforce ranking consistency. Second, a supervision mismatch arises because a censored patient's observed time indicates survival beyond that point and cannot serve as an exact regression target, although it still implies orderings relative to patients who died earlier. To address these mismatches, we propose CACSurv, a Concordance-Aligned Comparative framework for report-centric survival prediction. CACSurv reformulates survival modeling as mini-cohort comparative reasoning, where an LLM predicts relative prognostic orderings. We introduce concordance-aligned rewards derived from comparable relations under right censoring, enabling censored outcomes to provide ranking supervision without exact event-time targets. At inference, Monte Carlo Reference Aggregation compares each patient with sampled references and aggregates positions into a cohort-level ranking. We establish TCGA-SurvReport, a benchmark covering six TCGA cancer cohorts. CACSurv achieves the highest C-index on all six cohorts and an average C-index of 0.722, outperforming the strongest published survival model by 6.5 percentage points and the strongest LLM time-regression baseline by 4.2 percentage points. Our code, models, and dataset will be available at https://github.com/xmed-lab/CACSurv.

cs.AI↗

KHiM-Mamba: Injecting Pathology Knowledge into Mamba via Hidden-State Modulation for Whole Slide Image Analysis

Whole slide image analysis is commonly formulated as multiple instance learning (MIL), where instance features are contextually updated and aggregated into a slide representation, a process we term slide encoding dynamics. Recently, selective state-space models (SSM) have emerged as promising MIL architectures due to their long-sequence modeling capability and linear complexity. However, existing SSM-based MIL methods rely solely on visual features during MIL. Meanwhile, in large-scale WSIs, where sparse diagnostically decisive regions are surrounded by abundant irrelevant information, such purely vision-driven selective dynamics can misallocate state updates and readouts, causing the evolving SSM state to accumulate task-irrelevant evidence and dilute critical diagnostic cues over long scan trajectories. In this work, we propose the Knowledge-Aware Hidden-State Modulation architecture (KHiM-Mamba), which innovatively regulates Mamba's core selective state-space mechanism with explicit knowledge priors, steering slide encoding dynamics toward diagnostically meaningful evidence accumulation. Specifically, we redesign the original SSM layer to perform knowledge modulation operations during the evolution of hidden states, thereby guiding what visual evidence is accumulated and retrieved from the hidden state at each encoding step. Furthermore, we additionally introduce a local-adaptive vocabulary retrieval module that uses large language models to assign each patch fine-grained, tissue-specific semantic descriptions, enabling precise modulation across diverse tasks. Experiments on 11 public benchmarks across 4 tasks show that KHiM-Mamba consistently achieves state-of-the-art performance.

eess.IV↗

NARRATE: A Multimodal Real-World Australian Driving Dataset for Human-Centred Explanations in Automated Driving

Automated vehicles must explain their decisions in ways that passengers can understand, monitor, and trust. Existing language-annotated driving datasets are mostly observer-written, post-hoc, simulation-based, or generated from sensor inputs, rather than elicited from the driver performing the action. We introduce NARRATE, a multimodal real-world Australian driving dataset comprising 2,050 annotated events from 35 experienced drivers and driving instructors on public roads. Each event is grounded in synchronised visual, localisation, motion, and LiDAR streams and paired with in-vehicle and/or post-drive free-text explanations. NARRATE provides action labels, scenario-context labels spanning six high-level and 32 fine-grained categories, and span-level Situational Awareness (SA) annotations over driver explanations for Perception, Comprehension and Projection. Four benchmark tasks (SA, scenario-context, driver-action classification, and explanation generation) show that this structure is learnable from driver language, while fine-grained context recognition and explanation generation remain challenging. NARRATE paves a path towards more human-centred and domain-aware explanation models for automated driving.

cs.CV↗

Forgetting-Resistant and Lesion-Aware Source-Free Domain Adaptive Fundus Image Analysis with Vision-Language Model

Source-free domain adaptation (SFDA) aims to adapt a model trained in the source domain to perform well in the target domain, with only unlabeled target domain data and the source model. Taking into account that conventional SFDA methods are inevitably error-prone under domain shift, recently greater attention has been directed to SFDA assisted with off-the-shelf foundation models, e.g., vision-language (ViL) models. However, existing works of leveraging ViL models for SFDA confront two issues: (i) Although mutual information is exploited to consider the joint distribution between the predictions of ViL model and the target model, we argue that the forgetting of some superior predictions of the target model still occurs, as indicated by the decline of the accuracies of certain classes during adaptation; (ii) Prior research disregards the rich, fine-grained knowledge embedded in the ViL model, which offers detailed grounding for fundus image diagnosis. In this paper, we introduce a novel forgetting-resistant and lesion-aware (FRLA) method for SFDA of fundus image diagnosis with ViL model. Specifically, a forgetting-resistant adaptation module explicitly preserves the confident predictions of the target model, and a lesion-aware adaptation module yields patch-wise predictions from ViL model and employs them to help the target model be aware of the lesion areas and leverage the ViL model's fine-grained knowledge. Extensive experiments show that our method not only significantly outperforms the vision-language model, but also achieves consistent improvements over the state-of-the-art methods. Our code will be released.

cs.CV↗

Beyond Frame Selection: Generative Latent Evidence Aggregation for Long-Video Understanding

Long-video understanding commonly compresses videos into a small set of frames or visual tokens for answer generation. Existing compact pipelines focus on retaining relevant visual content as explicit evidence. Yet making evidence available does not ensure that complementary cues across moments are integrated for answering. Our key idea is to organize selected frames into query-relevant cross-frame evidence before generation. We formulate this post-selection stage as a latent evidence interface and instantiate it with GenEvA ($\textbf{Gen}erative$ $Latent$ $\textbf{Ev}idence$ $\textbf{A}ggregation$), a distribution-guided latent evidence aggregation framework. Specifically, GenEvA uses a query-conditioned evidence distribution to focus aggregation on relevant frames, forming compact cross-frame latent evidence from their frame-specific information. Since cross-frame integration is not always needed, the same distribution determines whether to insert this latent complement. Across four benchmarks and two Video-MLLM backbones, GenEvA consistently improves matched-frame baselines. At 8 frames, it raises the four-benchmark LLaVA-Video average by $+5.2$ points and Qwen2.5-VL accuracy on LVBench by $+10.1$ points. These gains require only $0.11\%$--$0.40\%$ average video-token overhead; analyses further show task-aware allocation and benefits from Adaptive Evidence Invocation.

cs.CV↗

SecRespond: Benchmarking AI Agents for Real-World Post-Compromise Incident Response

Large Language Model (LLM) agents are increasingly adopted in real-world security operations with access to host artifacts and command-line interfaces (CLIs), making it critical to thoroughly assess their security capabilities. However, existing cybersecurity benchmarks focus on pre-compromise settings where agents are placed in a clean and idealized environment before an attack occurs. This leaves the post-compromise setting underexplored. To address this gap, we introduce SecRespond, the first benchmark for evaluating LLM agents on the post-compromise incident-response workflow. Given a forensic disk snapshot of a compromised host together with the alerts, vulnerability scans, and baseline checks reported by a host security product, agents are required to produce forensic reports on intrusions, baseline risks, and vulnerability risks, together with a remediation plan. We instantiate this task across 10 cyber ranges, each constructed from a distinct compromised cloud host, spanning 4 entry-point types, 21 ATT&CK techniques, and 5 operating systems. We evaluate 23 frontier LLMs on the OpenCode agent harness. Experimental results show that although current agents can reliably uncover the problems exposed by alerts, they struggle to proactively investigate the disk for silent intrusions and to produce comprehensive, verified remediation plans, with no model achieving complete detection and remediation on any single range. This reveals a fundamental bottleneck in building agents for real-world incident response. The benchmark is publicly available at https://github.com/Alibaba-NLP/qqr/tree/main/data/secrespond.

cs.CR↗