MMBU: A Massive Multi-modal Biomedical Understanding Benchmark to Probe the Perception Capabilities of Vision-Language Models

Vision and language models (VLMs) hold immense promise to transform biomedical imaging workflows, from detecting lesions in chest X-rays to profiling cellular features in microscopy. Realizing this potential, however, requires robust and fine-grained visual perception. Models need to correctly interpret subtle features in images, and they must do so across diverse biomedical modalities, scales, and contexts. Nevertheless, current benchmarks remain limited. To address these gaps, we introduce the Massive Multimodal Biomedical Understanding (MMBU) benchmark. It is the largest biomedical vision and language benchmark to date, covering 35 submodalities with rich structured metadata. It includes both open and closed versions of ungrounded classification, grounded classification, and object detection, enabling systematic evaluation of model performance across biological scales, clinical settings, and imaging modalities. Evaluating 15 open-weight and 2 frontier VLMs, we find that while medical adaptation provides measurable gains for some models, the high accuracy often reported on established benchmarks can mask deficiencies in visual perception and domain generalization.

Beyond Output Matching: Preserving Internal Geometry in NVFP4 LLM Distillatio

Demand for low-precision inference, including NVFP4-based approaches, has grown as large language models are increasingly deployed in latency and cost constrained production environments. Quantization-aware distillation (QAD) helps recover accuracy lost under low bit quantization by training a quantized student to match the output distribution of a frozen higher precision teacher via a KL-divergence loss. In this work, we first provide a representation level diagnosis of QAD: output matching alone can mask internal degradation, because many intermediate activation geometries can yield similar teacher-aligned logits. Using CKA, we show that KL-only QAD can reduce layerwise representational similarity relative to the BF16 teacher, with especially severe drift in RL-post-trained models. This drift correlates with downstream bottlenecks on reasoning and coding tasks, suggesting that low bit recovery requires preserving internal geometry rather than matching outputs alone. Motivated by this finding, we propose \textbf{CKA-QAD}, a CKA-guided representational alignment method for NVFP4 QAD and low bit LLM accuracy recovery. The method adds a lightweight regularizer that preserves internal representational geometry during distillation by aligning layerwise Gram matrices through CKA. Across Nemotron 3 Nano and Qwen3-4B-Thinking-2507, CKA-QAD substantially improves representational alignment and improves downstream reasoning and coding accuracy with modest training overhead. Our findings position CKA-guided representational alignment as a practical complement to output matching for quantized LLM recovery.

Diffusion-Based Heart Sound Generation: Evaluation with Physiological Signal Metrics, Classifiers, and Expert Listening

Publicly available phonocardiogram (PCG) datasets remain limited in size and pathological diversity, constraining both auscultation training and the generalisation of automated heart-sound classifiers. A class-conditional diffusion model for PCG generation is developed in the log-mel domain and synthetic fidelity is assessed using complementary (i) physiology-inspired plausibility metrics, (ii) downstream label-consistency evaluation, and (iii) expert listening. Experiments use the Phy-sioNet/Computing in Cardiology Challenge 2016 dataset (3240 recordings) with recording-level splits. After preprocessing and quality control, 16,749 non-overlapping 4 s clips are mapped to a normalised 1 x 128 x 128 log-mel representation to train a conditional 2D U-Net denoiser with classifier-free guidance. Signal-level plausibility is quantified on reconstructed waveforms using three lightweight metrics: an envelope-autocorrelation rhythm score, an amplitude-based explosion score, and the dominant cycle lag. Synthetic clips preserve similar dominant cycle durations but exhibit reduced envelope periodicity and increased transient burstiness relative to real clips. For downstream evaluation, a ResNet-50 classifier achieves 92.24% accuracy on the held-out real test set and 82.8% accuracy on class-balanced synthetic batches, indicating that generated signals retain discriminative structure relevant to normal/abnormal classification. In a pilot expert listening study (60 clips, two clinicians), most synthetic clips are judged as heart-sound-like, while abnormality sensitivity is low for both real and synthetic 4 s excerpts. Overall, the results provide a practical baseline for diffusion-based PCG generation while highlighting remaining challenges in retaining abnormal acoustic cues and reducing reconstruction-induced artefacts.

IMAC-AgriVLN: Can Agricultural Vision-and-Language Navigation Agents be Aware of Instruction Mistakes?

Agricultural robots are serving as powerful assistants across a wide range of agricultural tasks, nevertheless, still heavily relying on manual operations or railway systems for movement. The AgriVLN method and the A2A benchmark pioneeringly extended Vision-and-Language Navigation (VLN) to the agricultural domain, enabling a robot to navigate to a target position following a natural language instruction. However, almost all the prior methods adopt an ideal assumption that the given instructions themselves are correct, which does not align with the realistic scenarios, because anybody may say an instruction with mistakes. To bridge this gap, we propose the A2A-MI benchmark, in which we build a semi-automatic data annotator to insert three mistake classifications into each original instruction in a more diversified and efficient way. We test several state-of-the-art agricultural VLN agents on it and observe a sufficient drop with -57% on SR and -9% on NE, from which we suggest that an agricultural VLN agent tends to assume that the given instruction is correct, so does not have the awareness to doubt it when the scenes it sees do not align with the instruction it receives. To build the awareness on instruction mistake, we propose the IMAC module analyzing the instruction and the current front-facing image, to judge whether the instruction has mistakes and attempt to correct it when needed. We integrate IMAC into the baseline model, and observe a noteworthy improvement, sufficiently narrowing the gap to the performance on instructions without mistakes. Project: https://github.com/AlexTraveling/IMAC-AgriVLN.

OmniMatBench: A Human-Calibrated Multimodal Reasoning Benchmark Across 19 Materials Science Subfields

As multimodal language models play an increasingly important role in scientific research, materials science offers a critical testbed due to its interdisciplinary, multimodal, and application-driven nature. However, existing materials benchmarks mainly focus on property prediction, knowledge QA, or characterization understanding, leaving the broader reasoning process from materials knowledge to application underexplored. To fill this gap, we present OmniMatBench, a human-calibrated multimodal reasoning benchmark for materials science. OmniMatBench contains 3,171 expert-curated QA and calculation problems across 19 materials-science subfields, spanning fundamental materials knowledge, structural and engineering materials, materials processing and manufacturing, and functional and applied materials. We evaluate 13 open-source and closed-source MLLMs and find that the best model achieves only a 0.372 overall score, revealing a substantial gap in current materials-science reasoning. Further analysis shows strong variation across subfields, fixed reasoning heuristics, uneven materials knowledge, and limited high-level knowledge application under formula-, retrieval-, and code-assisted settings. OmniMatBench provides crucial insights into the capabilities and limitations of current MLLMs and establishes a foundation for reliable AI assistants in materials-science research.

On-Policy Replay for Continual Supervised Fine-Tuning

Continual supervised fine-tuning (SFT) is the de facto recipe for adapting large language models (LLMs) to a stream of downstream tasks, but it suffers from catastrophic forgetting of earlier capabilities. Recent work shows that on-policy signals -- training on the model's own outputs -- reduce forgetting more reliably than off-policy supervision. Existing on-policy methods route this signal through a new training objective (e.g., self-distillation losses with a teacher copy), inheriting an extra forward pass, schedule sensitivity, and stylistic drift from the teacher.We instead route the on-policy signal through the training data source. Our method, On-Policy Replay (OPR), rolls out the most recent checkpoint on a small budget of historical prompts, filters the generations by a task reward, and replays the surviving (prompt, model response) pairs as ordinary SFT examples. There is no teacher, no auxiliary loss, and no on-the-fly distillation. Across three 7--8B instruction-tuned backbones (Qwen2.5-7B-Instruct, Qwen3-8B, Llama3.1-8B-Instruct) on the TRACE continual-learning benchmark, OPR consistently reduces forgetting; on the sharpest stress test (Qwen2.5-7B-Instruct, Sequential SFT BWT -13.93), OPR lifts BWT to -0.65 at a 10% replay budget and to -2.29 at a 1% budget -- a 46% reduction in |BWT| over a tuned Vanilla Replay baseline, with 42--46% reductions observed across all three backbones. We give a KL-shrinkage interpretation that places OPR and prior on-policy distillation methods on a single axis, and we present a counterintuitive finding that explains why Vanilla Replay is already a strong baseline: low-score replay is uniformly worse than Vanilla Replay, demonstrating that the active ingredient in OPR is the on-policy distribution, not the response quality alone.Our code is available at https://github.com/Yancey2024/OnPolicyReplay.

How Far Has AI Come in Liver Fibrosis Staging? A Large-Scale Real-World Dataset and Benchmark

Despite years of methodological progress, how far AI has come in liver fibrosis staging has never been systematically evaluated under the heterogeneous, multi-center conditions that define clinical practice. To address this gap, we introduce LiFS, a large-scale dataset and benchmark derived from the MICCAI 2025 CARE-Liver challenge, comprising 610 patients across multiple centers and scanners with multi-sequence MRI. To the best of our knowledge, LiFS is the first benchmark providing complete gadoxetic acid-enhanced sequences with histopathology-confirmed annotations from diverse real-world scanners. Through systematic evaluation of 9 independently developed methods selected from 96 registered teams against in-cohort radiologist reference results, our findings address how far current AI has progressed toward clinical-level liver fibrosis staging from three complementary perspectives. First, against radiologists, the best AI methods were broadly comparable to the senior radiologist and significantly exceeded the junior radiologist in selected settings, while median AI performance generally approached junior-radiologist levels. Second, from a data perspective, cross-center heterogeneity, label imbalance, and contrast-enhanced sequence variability emerge as the dominant challenges for AI methods. Third, from a technical perspective, methodological design choices, including spatial registration, input dimensionality, multi-modal fusion strategy, and backbone architecture, appear to modulate cross-center robustness, although no single choice alone closes the gap. Overall, LiFS provides a rigorous real-world benchmark for positioning the current state of AI in liver fibrosis staging and for enabling future research on the key challenges that limit clinically reliable deployment.

MIND: Multi-Scale Intent Diffusion for Text-Driven Physics-Based Humanoid Control

Enabling physics-based humanoids to execute diverse behaviors from high-level textual commands remains a significant challenge. Existing methods typically follow either a two-stage paradigm that combines kinematic motion generation with physics-based tracking, or an end-to-end imitation-learning paradigm that directly generates actions from text. However, the former suffers from the inherent domain shift between kinematic generation and physics-based tracking, while the latter struggles with the substantial modality gap between textual commands and low-level actions, limiting effective semantic alignment. Notably, humanoid states encode rich motion dynamics that are more semantically aligned with textual descriptions than low-level actions, making them a natural basis for deriving behavioral intent. Building upon this insight, we propose MIND, a novel end-to-end diffusion framework for text-driven physics-based humanoid control that leverages behavioral intent as a semantic bridge between textual commands and low-level actions. At its core, MIND introduces a multi-scale intent diffusion mechanism, where a holistic intent predictor captures global behavioral dynamics to guide overall behavior synthesis, while an immediate intent predictor provides step-wise, fine-grained signals for local behavior refinement at each diffusion step. This hierarchical intent formulation imposes a structured inductive bias for humanoid control, improving semantic alignment and behavioral naturalness. Furthermore, MIND encodes humanoid states into a latent space to enable more effective semantic intent modeling. Extensive experiments demonstrate that MIND outperforms existing methods and synthesizes coherent, physically plausible, and semantically aligned humanoid behaviors from text commands. Our code will be released to facilitate future research.

LANG: Reinforcement Learning for Multilingual Reasoning with Language-Adaptive Hint Guidance

Reinforcement learning has proven effective for enhancing multi-step reasoning in large language models (LLMs), yet its benefits have not fully translated to multilingual contexts. Existing methods struggle with a fundamental trade-off: prioritizing input-language consistency severely hampers reasoning quality, while prioritizing reasoning often leads to unintended language drift toward English. We address this challenge with LANG, a novel framework that leverages language-conditioned hints to guide exploration in non-English reasoning tasks. Our method incorporates two key mechanisms to prevent dependency on these hints: a progressive decay schedule that gradually withdraws scaffolding, and a language-adaptive switch that tailors learning horizons to specific language difficulties. Empirical results on challenging multilingual mathematical benchmarks reveal that LANG substantially enhances reasoning performance without compromising language consistency. Moreover, we show that our framework generalizes beyond mathematics, fostering more consistent language alignment across model layers

MTR-Suite: A Framework for Evaluating and Synthesizing Conversational Retrieval Benchmarks

Accurate evaluation of conversational retrieval is pivotal for advancing Retrieval-Augmented Generation (RAG) systems. However, existing conversational retrieval benchmarks suffer from costly, sparse human annotation or rigid, unnatural automated heuristics. To address these challenges, we introduce MTR-Suite, a unified framework for auditing, synthesizing, and benchmarking retrieval. It features: (1) MTR-Eval, an LLM-based auditor quantifying alignment gaps in previous benchmarks; (2) MTR-Pipeline, a multi-agent system using greedy traversal clustering to generate high-fidelity dialogues at 1/400th human cost; and (3) MTR-Bench, a rigorous general-domain benchmark. MTR-Bench mimics production-style challenges (hard topic switching, verbosity), offering superior discriminative power. We make our code and data publicly available to facilitate future research at https://github.com/rangehow/mtr-suite.