CATF: A Manager-Worker Framework for Context-Aware Time-Series Forecasting
Yueqiao Chen, Dongyu Liu
AI 速览：提出一种基于 Manager-Worker 架构的上下文感知时间序列预测框架，用 Transformer 经理按上下文硬路由到单个专家工人，使任意单模型预测器无需人工划分数据即可实现专业化。
We study context-aware time-series forecasting, in which a single target variable is predicted from its own history and a set of contextual variables, such as weather and calendar information. As the context becomes more diverse, a single model struggles to capture all of its regimes, while ensemble learning and Mixture-of-Experts approaches rely on manual data partitioning or soft routing, which increases the inference cost and hinders interpretability. We propose CATF, a Context-Aware Time-Series Forecasting framework built on a Manager-Worker architecture: a Transformer-based Manager routes each input to a single Worker based on the context, using hard routing during both training and inference. The Workers thus specialize without manual data partitioning, and any single-model forecaster can serve as a Worker. On five context-rich datasets, four horizons, and two normalization schemes, CATF improves over its standalone counterpart in 162 of 200 comparisons, the best CATF-enhanced model outperforms the best standalone baseline in 19 of 20 settings, and it outperforms the zero-shot Time-MoE foundation model in 15 of 20 settings. A case study on bike-sharing data shows that the Manager routes inputs by the calendar and that each Worker specializes in a different weekly pattern.
AI 速览：作者提出首个面向多模态有状态交互的大规模多轮对话数据集 YANAMI-400K 与评测基准 YANAMI-Bench，并报告微调 BAGEL 后在图像编辑、生成与有状态交互上均有提升。
Unified Multimodal Models (UMMs) integrate visual understanding, generation, and editing within a single unified framework. However, existing unified multimodal datasets and benchmarks remain dominated by single-turn tasks. Even in recent multi-turn datasets, many turns can be solved independently without tracking information across turns. Consequently, they provide limited supervision and evaluation for multimodal stateful interaction, which requires models to understand, maintain, update, and faithfully realize evolving dialogue states across turns. To address this gap, we introduce YANAMI, the first large-scale multimodal dialogue dataset and benchmark for stateful interaction, comprising YANAMI-400K and YANAMI-Bench. YANAMI-400K contains 400K multi-turn multimodal dialogues, with more than 2M images and 2.6M dialogue turns. The dialogues span four representative scenarios: Multi-Image Fusion, Image Editing, World Modeling, and Visual Reasoning, and require models to continuously track, preserve, and revise evolving multimodal dialogue states across turns. YANAMI-Bench comprises 400 multi-turn dialogues for systematically evaluating multimodal stateful interaction. It jointly evaluates Dialogue State Understanding and Visual Realization, characterizing the gap between models' ability to understand evolving dialogue states and their ability to faithfully realize them in visual outputs. Experiments show that fine-tuning BAGEL on YANAMI-400K improves image editing, generation, and multimodal stateful interaction by 10.3%, 2.0%, and 13.23%, respectively, while preserving its general visual understanding capabilities. Evaluations on YANAMI-Bench further reveal that existing UMMs still struggle to maintain, update, and visually realize evolving multimodal dialogue states.
iGC: Antibody Optimization via Reinforcement Learning over a Learned Somatic Hypermutation Prior
Thao Nguyen, Robert G. Alberstein, Ewa Nowara, Homa MohammadiPeyhani, Holly R. Steach, Sarah A. Robinson, Yuval Shechter, Donald W. Lee 等 13 位作者
AI 速览：iGC 是一个两阶段强化学习框架，先学习体细胞超突变先验，再在抗原特异性亲和力优化中结合 KL 正则与奖励税元梯度，以兼顾生物学合理的突变与结合优化。
Antibody affinity maturation is a natural optimization process in which B cells acquire mutations through somatic hypermutation (SHM), undergo selection for improved antigen binding, and iteratively evolve toward higher affinity. Computationally modeling this process requires capturing both biologically plausible mutations and antigen-specific selection. However, existing mutation models do not directly optimize binding, while affinity optimization often neglects realistic mutation trajectories. We introduce iGC (in silico Germinal Centers), a two-stage reinforcement learning (RL) framework for antigen-specific antibody optimization. Stage 1 learns an SHM mutation prior from parent-child pairs in B-cell lineage trees and adapts it to a target affinity-maturation dataset. Stage 2 initializes an RL policy from this prior and optimizes antigen binding with Kullback-Leibler (KL) regularization toward the frozen prior. To prevent reward exploitation, including repetitive high-reward substitutions and structurally tolerated framework mutations, we introduce a Reward-Tax Meta-Gradient objective. RewardNet smooths the affinity landscape, while TaxNet penalizes reward overexploitation, jointly defining a learned meta-reward for policy optimization. We evaluate iGC on mutation prediction, SHM consistency, affinity optimization, and mutation diversity. The mutation prior more than doubles the strongest baseline in mutation-site prediction. Following RL optimization, iGC maintains SHM sequence structure. It also achieves a strong exploration-exploitation balance, yielding competitive reward-diversity products across candidate-selection cutoffs. These results demonstrate that iGC integrates biologically plausible mutation generation with antigen-specific affinity selection, providing an in silico framework for iterative antibody affinity maturation.
Establishing almost sure convergence rates for stochastic approximation and reinforcement learning under Markovian noise is a fundamental theoretical challenge. We make progress towards this challenge for a class of stochastic approximation algorithms whose expected updates are contractive, a setting that arises in many reinforcement learning algorithms such as $Q$-learning and linear temporal difference learning. Specifically, for a power-law learning rate $O(n^{-η})$ with $η\in (1/2, 1)$, we obtain an almost sure convergence rate arbitrarily close to $o(n^{1 - 2η})$. For the harmonic learning rate $α_n=α/(n+1)$, the almost sure squared-error rate is $o(n^{-ζ})$ for every $ζ<\min\{1,2(1-κ)α\}$, where $κ$ is the contraction factor. Thus exponents arbitrarily close to one are available when $2(1-κ)α\geq1$. This approaches the $n^{-1}$ polynomial exponent of classical law-of-the-iterated-logarithm benchmarks. Key to our analysis is a novel Lyapunov drift construction that applies a Poisson-equation based correction for Markovian noise to the well-established Moreau-envelope smoothing for the contractive mapping.
Video diffusion alignment has been heavily relied on scalar rewards. These rewards are typically derived from learned reward models in human preference datasets, requiring additional training and extensive collection. Moreover, scalar rewards provide coarse, global supervision, offering limited prompt-generation mismatch credit assignment and making models prone to reward exploitation and unstable optimization. We propose Diffusion-DRF, a free, rich, and differentiable reward framework for video diffusion fine-tuning. Diffusion-DRF employs a frozen, off-the-shelf Vision-Language Model (VLM) as the critic, eliminating the need for reward model training. Instead of relying on a single scalar reward, it decomposes each user prompt into multi-dimensional questions with freeform dense VQA explanation queries, yielding information-rich feedback. By direct differentiable optimization over this rich feedback, Diffusion-DRF achieves stable reward-based tuning without preference datasets collection. Diffusion-DRF achieves significant gains both quantitatively and qualitatively, outperforming state-of-the-art Flow-GRPO by 4.74% in overall performance on unseen VBench-2.0.
The ecosystem of Lean and Mathlib has become the de facto standard for large language model (LLM) assisted formal reasoning with remarkable successes in recent years. Those successes, however, only consume Mathlib as an essential dependency but do not directly contribute to it. In the meantime, the growth of Mathlib has recently been bottlenecked by the review process, which requires human reviewers to judge whether proposed pull requests (PRs) follow the Mathlib's conventions and are worth integrating as part of a shared mathematical infrastructure. This leads to our central question: can LLMs help review Mathlib PRs? To this end, we introduce MathlibPR, a benchmark built from real Mathlib4 PR histories. We further propose a staged evaluation protocol and use it to evaluate both LLM models (e.g., DeepSeek, Qwen, Goedel, and Kimina) and LLM agents (e.g., Codex and Claude Code). Surprisingly, both LLM models and LLM agents struggle to distinguish merge-ready PRs from build-passing PRs that were revised or never merged. By turning Mathlib PR histories into a supervised signal, MathlibPR provides a step toward reviewer assistants and reward models that could help evaluate PRs and steer LLMs toward producing merge-ready Mathlib contributions.
LLM agents are judged by what they deliver, and in professional work what gets delivered is a project, not a task. A project pays off only if the whole package holds, and one weak part can limit the value of the rest. Yet existing benchmarks mostly test agents on one task at a time, leaving a gap between the capability we measure and the outcome that makes agents useful. To address this gap, we introduce DeliverBench, which tests whether agents can carry a project from brief to final delivery. Its 11 projects, spanning financial due diligence, legal contract drafting, and drug safety reporting, are reconstructed with 13 domain experts from real materials and chain 150 subtasks through 2-15 rounds of new materials and client feedback. Because no reference outputs are injected, agents inherit their own errors from one task to the next. Expert rubrics then score every checkpoint as well as the final package. Across seven model-harness systems, even the best system scores only 65.0 out of 100 on final delivery and passes 50.0% of subtasks at a 90-point threshold. The failures are often failures of carrying work forward: in one audit project, all seven systems addressed the review findings, but only one preserved traceability. Together, these results suggest that task-level capability does not simply compose into delivery, which also depends on keeping the work coherent as it evolves.
Video generation has progressed from isolated clips to coherent short-form audiovisual sequences, and the bottleneck is shifting from rendering to semantic control. This report summarizes the observations and insights on multimodal understanding gained from successive generations of Kling. We argue that generation quality is the projection of multimodal understanding through the entire data--conditioning--alignment--evaluation chain, and we examine this chain module by module. Kling-MLLM, a shared multimodal foundation model that unifies image, video, audio, synchronized audio--video, and interleaved inputs, is specialized for each downstream module. Because a DiT is trained on far fewer training samples than a language model, we treat captioning as a semantic contract that maximizes description density, and extend it to reference-grounded captioning that makes reference bindings explicit. Prompt enhancement translates user intent into this caption distribution: basic PE is derived from captions of compelling videos and refined with text-level and generator-in-the-loop reinforcement learning, while agentic PE refines the story and retrieves compelling examples. We show that a reliable text encoder comes from perceptual training rather than from stronger language ability, and that candidates can be screened from text alone before end-to-end training. Reward modeling pairs a general reward model for explainable, localized critiques with a decision reward model for fast multidimensional scoring, treating perceptual bandwidth as a first-class design parameter. Finally, we review task-specific automatic evaluation, focusing on protocols that expose scoring shortcuts and help trace failures to their upstream source. We hope these lessons offer practical guidance for building the understanding stack behind controllable video generation.
LLM agents choose tools and arguments from context that mixes user requests, tool outputs, retrieved records, memory, and untrusted text. Evidence can be relevant without being authorized to determine a decision, so a correct action need not be grounded only in permitted evidence. We introduce a target-specific authorization audit that labels context factors separately for each tool and argument target. Its primary test holds the task, proposition, position, and policy fixed while changing only the proposition's source authority. We then test behavior when valid evidence is weakened and use context-subset interactions as a secondary localization diagnostic. Across 450 controlled next-action tasks and multiple open-weight LLM families, trusted and untrusted variants produce different actions in 5.4 percent of competing cases versus 1.7 percent of supporting cases. Under controlled degradation, unauthorized competition is retained in a full-correct, mixed-error, clean-correct pattern in 2.4 percent of comparisons, with a 95 percent confidence interval from 2.1 to 3.0 percent. These are controlled stress-set rates, not deployment prevalence. The models respond to textual source-authority cues, but this does not prevent untrusted evidence from influencing their actions.
Large language models (LLMs) are increasingly expected to extract and synthesize structured information from complex real-world documents, yet they typically produce unstructured, hard-to-verify paragraph-style outputs. We introduce AOE (Arranged and Organized Extraction), a bilingual benchmark that evaluates LLMs' ability to distill long-form, multi-document inputs (64k to 1.8M tokens) into structured tables. AOE comprises 12 tasks across four domains (academic, financial, legal, and medical), each grounded in genuine analytical needs, with schema guidance provided to isolate extraction quality from formatting variance. We evaluate 12 frontier LLMs (both open-weight and proprietary) under four generation settings of increasing autonomy: direct full-document input, RAG-based retrieval, offline document-browsing agent, and online web-search agent. Even the best-performing model achieves only 42.7% Cell F1, and retrieval-augmented approaches often degrade rather than improve performance, revealing that structured extraction from long contexts remains a fundamental challenge - an "Area of Effect" (AOE) that impacts all tested systems.
Training neural operators on million-cell meshes can exceed GPU memory limits. Models are therefore trained on subsamples and deployed on whole meshes or selected regions. However, subsampling can distort boundary information, degrading accuracy. We ask when boundary conditioning under sampling stays consistent with whole-mesh training, and how to aggregate and inject boundary data. Representing the solution as an integral of boundary data against type-dependent measures, we show that subsampling distorts this conditioning in three ways: incomplete boundary coverage or a sample-dependent global state, face weighting by count rather than area, and compression of boundary inputs into a low-dimensional vector. This analysis yields four measurable consistency conditions checkable before training. We satisfy them with two backbone-agnostic modules. Boundary Quadrature Attention (BQA) aggregates all boundary faces, typed and area-weighted, into slice tokens that every query point reads at every layer. Anchored Global State (AGS) estimates the global state from anchors independent of the sampled points. Adding both to a scalable message-passing and slice-attention backbone yields SCANO. We evaluate the approach on controlled and public benchmarks and a data-centre thermal simulation dataset (DCTS) containing million-cell meshes. On both benchmarks, BQA improves every backbone, reducing whole-mesh error up to 15-fold. On all three datasets, AGS aligns regional with whole-mesh predictions, cutting regional error at least tenfold for most backbones on the public benchmark. Together, the modules nearly halve the best original backbone's whole-mesh error on both benchmarks. SCANO predicts a room-scale field in under one second, over 100 times faster than whole-room anchor self-attention.
Joint nuclei instance segmentation and classification requires precise boundaries and neighborhood information that helps distinguish morphologically similar cell types. We present TME-ContextNet, a multi-task framework that learns class-conditioned spatial context from existing nuclear annotations and organizes decoding into geometry-related and semantic task groups. Class-conditioned spatial supervision extends nuclear labels into a dense target using Euclidean distance transforms, requiring no additional tissue-region annotation. We derive its nearest-region interpretation and characterize its stability to small annotation perturbations. A local--global context module combines dilated convolution and self-attention, while a grouped task decoder reduces redundant computation. The spatial head supplies training supervision and is omitted during inference. Experiments on PanMix, CoNSeP, and PanNuke yield Dice scores of 80.23%, 65.32%, and 79.13%, and mean multi-class panoptic quality (mPQ+) scores of 58.67%, 35.50%, and 43.74%, respectively. PanNuke results are averaged across three test folds. Target and decoder controls support the combination of distance-based supervision and grouped decoding while identifying their dependence on task configuration. The framework uses 36.40 million parameters, compared with 54.74 million for HoVer-Net. These findings support a compact, context-aware formulation for joint nuclei analysis in computational pathology.
Audio language models are meant to answer questions from the audio itself, but an hour-long recording rarely fits in their context, and the answer often depends on a few seconds of it. Transcribing the recording first makes it fit but discards sounds and voices, while systems that return to the audio must be trained for this behavior. We propose ReListen, a training-free method in which a frozen audio language model finds its own evidence before answering. In a Jev-style scan, the model scores each 20-second window by the probability that it helps answer the question, and it then answers from the highest-scoring minute. This minute overlaps the annotated evidence far more often than transcript retrieval, for 88.5% of VoiceGiraffe questions. Without any change, the same scan also grounds events on LAT-Bench as well as a model trained for temporal grounding. Answering from it matches or exceeds hearing the whole recording with a small fraction of the audio, and on VoiceGiraffe it exceeds a transcript of the whole recording by 9.2 points. The selected evidence is also worth hearing rather than reading. Given the same excerpts, reading their transcript matches listening on speech content, but listening is clearly more accurate on questions about sounds and voices. Moreover, ReListen answers faster than every transcript-based system, and its memory stays constant even for recordings too long to hear whole on one GPU. Therefore, a frozen audio language model can search hour-long recordings by itself, without a transcript or additional training.
Syndrome extraction for a CSS code has depth at least the Tanner-graph degree $Δ$, attained by edge colouring if gate order is ignored; correctness adds a parity constraint. We prove three general obstructions to depth-$Δ$ extraction and show that, for every two-block group-algebra code with weight-three polynomials whose checks overlap in at most two qubits, no correct single-round circuit with one ancilla per check and one CNOT per Tanner edge has depth six: the minimum is seven, so the seven-layer cycle of Bravyi \emph{et al.} is optimal for their seven bivariate bicycle codes. This is the first analytic, family-wide impossibility result, extending prior instance-specific solver-based infeasibility results. The hypothesis cannot be dropped: a single orbit of four-qubit overlaps with a collision element of even order and $2$-height at most eight makes depth six attainable, and a $[[72,8,6]]$ code of this kind has a verified six-layer circuit. The refutations carry DRAT certificates. As a further application of the same scheduling scheme, we construct Stab-QRAM, a quantum random access memory for affine Boolean data that uses only CNOT and $X$ gates and has the minimum depth possible for this gate set.
Among 291 surveyed open-weight language model repositories (280 distinct embedding tables), 52.6% of untied gated and hybrid checkpoints (20/38) contain reachable input embeddings with norms below $10^{-6}$, compared with 0% of weight-tied models. Such near-zero rows can arise during pretraining when unseen rare or template tokens receive zero input gradients while AdamW weight decay shrinks untied embeddings. In tested Transformer configurations, non-initial positions or a BOS token prevent this failure mode. In gated sequence models (Mamba, Gated DeltaNet, xLSTM), a near-zero token can amplify input-embedding gradients by more than 30 orders of magnitude (up to 35). At zero input, current-token readout factors (SiLU gates or linear queries) vanish, suppressing historical-state readout. Residual RMS values can then fall below the square root of the RMSNorm floor across consecutive layers, causing Pre-RMSNorm to amplify backward gradients by up to $10^3$ per layer. When a gradient entry exceeds $1.84 \times 10^{19}$, FP32 global-norm clipping in PyTorch, Hugging Face Trainer, and FSDP2 can overflow to infinity and zero the current batch's gradients. In Telugu continual pretraining of Nemotron-3-Nano-4B, this discards updates on approximately 20% of steps and leaves triggered rows frozen. We propose two remedies. For released checkpoints, GroupClip decouples input-embedding clipping using a 64-bit or rescaled norm; its rescaled implementation eliminates update loss and heals triggered rows in one step with negligible measured end-to-end overhead. For pretraining, GateFloor ($\mathrm{SiLU}(z) + δ$, $δ\in [0.01, 0.1]$) keeps the gate positive at zero input, reducing near-zero gradients by 28 orders of magnitude in experiments up to 370M parameters, with perplexity comparable to the baseline.
As streaming generation becomes the dominant serving paradigm for both text-only and multimodal large language models, ensuring safety throughout the generation process has become a critical challenge. We introduce PGuard, a native, streaming, and plug-in guardrail that transforms safety monitoring from an external serving dependency into an integrated capability of the backbone model. PGuard directly leverages the backbone's intermediate representations for token-level safety detection, without altering its original inference pathway or generation behavior. Specifically, dynamic layer gating selectively aggregates residual-stream representations across layers to extract safety-relevant signals during both prefill and autoregressive decoding. To balance detection accuracy and computational efficiency, PGuard supports multiple representation-modeling architectures, including a lightweight stateless variant and sequence-aware variants built on multi-scale causal EMA-attention and GRU-based recurrent modules. Its design also accounts for the execution characteristics of LLM serving frameworks, such as SGLang, optimizing computation and data access to minimize additional runtime and synchronization overhead. As a result, PGuard incurs only 0.82-2.26% end-to-end overhead on the hundred-billion-parameter backbone Qwen3.5-122B-A10B served by SGLang. Extensive evaluations across more than 30 benchmarks spanning text and multimodal safety, jailbreaks, cyber and CBRN risks, multilingual scenarios, and real-world traffic demonstrate state-of-the-art performance among native guardrails. On text queries with Qwen3.5-9B, PGuard achieves an average F1 score of 0.836, approaching dedicated external guardrails, while supporting deployment-time configuration of category-specific thresholds and enforcement actions.
Wanshu Fan, Zheng Tong, Yang Liu, Jing Qin, Zhongbin Han, Haifan Gong, Congyu Liao, Xiaofeng Liu 等 9 位作者
Large language models and multimodal foundation models are enabling medical artificial intelligence (AI) systems to move beyond isolated prediction and undertake multistep clinical tasks that require planning, tool use, memory, iterative correction, and coordination among specialized agents. However, the scope of agentic AI in medicine remains unsettled, and current evaluation practices are not yet aligned with the requirements of clinical use. We conducted a scoping review with systematic evidence mapping across five electronic sources, screened 1,649 exportable records, and provisionally included 557 unique studies that met predefined criteria for goal-directed task execution, tool use, interaction with external resources, feedback-based refinement, or multi-agent collaboration. The included studies describe single agents that use external tools, workflows supported by retrieval and external knowledge, multimodal agents, and multi-agent systems applied to medical question answering, image interpretation, electronic health record analysis, drug safety, and clinical trial prediction. The evidence base remains dominated by public benchmarks, simulated settings, retrospective datasets, and small-scale expert evaluation. Process reliability, evidence traceability, uncertainty, safety, workflow impact, and external validity are evaluated less consistently. Clinical translation will depend on clearer definitions, reproducible evaluation, auditable oversight, interoperable system design, and prospective validation in real-world clinical workflows.
Medical world models offer a framework for extending medical artificial intelligence beyond static prediction by representing evolving patient states and modelling how they change over time and in response to clinical interventions. This Review uses structured narrative synthesis and reproducible evidence mapping to define conceptual boundaries, technical foundations, application domains, and clinical evidence requirements. We screened 1,455 unique records and assembled a corpus of 300 sources, including 14 studies that met a strict empirical definition of a medical world model. The field is organized around four capabilities: patient state representation, temporal dynamics modelling, intervention-conditioned simulation, and clinician-supervised planning. Evidence spans medical imaging, longitudinal electronic health records, treatment response modelling, physiological and multimodal state modelling, ultrasound and surgical interaction, and population and health-system simulation; clinical digital twins are treated as a cross-cutting integration framework. Studies provide early evidence of technical feasibility for trajectory forecasting and comparison of candidate interventions, but most remain retrospective, task-specific, or preclinical. We distinguish functional capability from clinical evidence maturity and specify evidence needed to support progressively stronger clinical claims, from temporal forecasting to intervention comparison and planning-oriented decision support. These requirements include precise intervention representations, robust causal and mechanistic grounding, calibrated trajectory-level uncertainty, safety-constrained planning, external validation, and prospective multicentre evaluation against clinically meaningful endpoints. This framework aims to distinguish technically plausible simulation from clinically supportable prediction and decision support.
Xiangyu Li, Jincheng Li, Xing Xie, Bin Guan, Shaohua Kevin Zhou, Rui Yan, Ran Su
Whole-slide multiple instance learning can predict diagnostic and prognostic labels from tissue composition while making limited use of spatial organization. Providing coordinates or a graph does not establish that a trained predictor uses tissue neighborhoods. We examine this gap as spatial blindness and hypothesize that early composition learning can weaken supervision of a spatial pathway. We introduce ResTopoMIL, which freezes a learned composition predictor before training a graph residual with a coordinate-reassignment separation loss. Permutation and information analyses distinguish architectural symmetry, available spatial information and the proposed optimization mechanism. On a controlled spatial task, staged training achieved an AUC of 0.987, compared with 0.684 for the same jointly trained architecture. Performance remained high on a composition-only task. Across four WSI classification and five survival benchmarks, ResTopoMIL improved mean primary performance on eight tasks; four improvements remained significant after Holm-adjusted patient-cluster tests. Coordinate-free residual controls recovered some of the gains, while neighborhood destruction reduced performance close to the composition anchor. Neighborhood-preserving transformations and anatomy-stratified perturbations further characterized this dependence. These results support fixing a composition reference as a practical intervention for learning spatial corrections, while retaining additional appearance capacity as a contributing explanation.
World models allow agents to anticipate action consequences, but proactive assistance also requires interpreting incomplete observations and deciding whether to intervene before a user asks for help. We propose See-Foresee-Act and instantiate it as the Proactive Intent World Model (PIWM) for retail assistance. See preserves role-relevant observable evidence alongside AIDA-BDI customer states; Foresee predicts one-step consequences for candidate actions; Act explicitly compares those consequences, including non-intervention. GuidanceSalesBench supplies predominantly synthetic supervision linking observations, states, candidate outcomes, and decisions. On 509 Full test examples, consequence-based selection achieves 63.65% decision accuracy, compared with 45.38% for direct decision output, an improvement of 18.27 percentage points. These results support candidate-consequence comparison for decision reasoning in this retail task. Code and data will be released after consent verification and upon publication.
Chest radiography (CXR) diagnosis is critical for thoracic disease assessment, yet challenged by the need to identify subtle lesions and co-occurring abnormalities across regions. While recent vision-language models can actively observe local regions, existing preference-based methods supervise each observation by its immediate effect on the current diagnosis, failing to capture how each observation and stopping decision contributes to the final diagnosis. To address this issue, we propose GroundSeek-CXR, an active CXR diagnosis framework that learns observation and stopping decisions by retrospectively evaluating them with the final diagnosis. Specifically, we first devise a Retrospective Diagnostic Evaluation (RDE) that scores observation and stop candidates from offline Monte Carlo tree search through independent continuations. Moreover, we design a Retrospective Gap-weighted Preference Optimization (RGPO) that learns from retrospective-score preferences weighted by their score gaps, prioritizing actions that improve diagnostic accuracy. Extensive experiments demonstrate that our GroundSeek-CXR framework outperforms state-of-the-art methods, and improves abnormal-image sample-F1 by 4.48% over immediate-score preference learning.
Synchrotron detectors generate data at rates that challenge storage and I/O systems. Compression can reduce data movement, but codec-only measurements do not capture the end-to-end behavior of HDF5 workflows. We present PhotonZip, a lifecycle-aware compression framework that tests candidate modes through a common HDF5 path and selects among eligible measured configurations according to either compression ratio or round-trip throughput. Its modular filters preserve standard HDF5 access when the required plugin is available. Across four beamline-derived datasets, PhotonZip reaches lossless compression ratios up to 1.629$\times$ and HDF5 write throughput up to 3.40GB/s write and 7.75GB/s read. The throughput-oriented configuration achieves 1.529$\times$ lossless data reduction while improving HDF5 write and read throughput by 21.2\% and 15.0\%, respectively. Lossy preview modes provide up to 16$\times$ data reduction and reduce preview latency by up to 8.35~s in the evaluated file-per-rank HDF5 configuration.
Ziyu Liu, Jingkai Hong, Robert S. Aviles, Sasan Razmkhah, Peter A. Beerel, Massoud Pedram
Rapid Single Flux Quantum (RSFQ) logic can operate at ultra-high clock frequencies; however, its physical design is complicated by explicit splitter trees and deep gate-level pipelines in which every logic gate is clocked. To address this challenge, we present qLevelPlace, a level-based placement framework for RSFQ circuits. We separate gate and splitter placement into two stages to better support levelguided gate-to-row assignment and use a timing-aware quadratic cost function to penalize long interconnects. Experiments on the benchmark circuits show that, on average, qLevelPlace achieves a $2.65\times$ higher clock frequency and a 51% reduction in layout area compared with the state-of-the-art RSFQ placer qPlace.
Louis Bradshaw, Irmak Bukey, Daniel Chenyu Lin, John Thickstun, Chris Donahue
Multimodal language models continue to improve on broad music understanding benchmarks, but which musical features can they perceive and reason about? We investigate this question by introducing PianoBench, a benchmark that tasks language models with reasoning about musical performances across congruent acoustic and symbolic input representations. PianoBench is derived from our new multimodal, training-scale dataset comprising more than 20,000 hours of interleaved transcripts of music and spoken musical commentary from instructional piano videos. Open-weight audio-language models achieve at most 52.8% accuracy on PianoBench, and at best only modestly outperform their accuracy on a no-audio control. In contrast, frontier models using chain of thought achieve up to 93.2% accuracy when provided transcripts of music as text, exposing a gap between perceiving musical features from audio and reasoning about musical events when they are given explicitly. We then test whether mid-training on interleaved musical instruction can improve question answering without chain of thought, raising Qwen 3.5 9B Base's zero-shot accuracy on PianoBench with music transcripts from 42.8% to 56.0% after instruction tuning. These experiments suggest that instructional videos may represent an overlooked source of multimodal supervision for music. Available: https://gclef-cmu.org/pianobench.
Tengfei Liu, Yang Shi, Yuran Wang, Yuqi Tang, Yizhuo Jia, Zhuoran Zhang, Qixun Wang, Shujie Li 等 15 位作者
Multimodal large language models (MLLMs) can generate detailed video captions, yet they may overemphasize minor details while overlooking essential context and key events. We argue that not all visual evidence deserves equal attention, as the importance of visual information depends on the content of each video. Motivated by this observation, we propose Reviso, an agentic video captioning framework that learns what to observe and how to gather relevant visual evidence. From a low-frame-rate global preview, Reviso constructs a hierarchical Context--Focus--Points (CFP) observation plan that identifies the video context, prioritizes key aspects, and defines concrete observation targets and corresponding tools. Guided by this plan, Reviso performs goal-directed active perception by selectively revisiting temporal segments and combining tools for fine-grained inspection. It interprets the acquired evidence, updates its plan as needed, and integrates global and local information to generate the final caption. To learn this observation strategy, we develop a verification-guided pipeline that constructs high-quality training trajectories through factual verification and teacher-guided refinement. We further train Reviso using Group Relative Policy Optimization (GRPO), incorporating a key-fact reward for caption quality, a Plan Alignment Reward (PAR) for identifying important observation targets, and an Unproductive Tool-Call Penalty (UTCP) that adjusts observation-cost penalties based on caption quality. Experiments across four video captioning benchmarks show that Reviso outperforms the evaluated open-source baselines, while its generated captions better support downstream video question answering. These findings highlight the effectiveness and potential of priority-guided active observation for video captioning.
Generative recommendation generates the semantic ID (SID) of the next item from a user's interaction history. Existing training objectives mainly supervise target SID tokens, while on-policy distillation (OPD) provides denser supervision through a teacher's probability distribution over all tokens for more effective and efficient training. However, OPD can aggravate popularity bias: teacher feedback favors popular-item tokens, transferring this preference to the student, while student rollout favors popular items, leaving unpopular-item SID paths rarely generated and supervised. These two sources of bias reinforce each other and reduce recommendation diversity. We propose HERO, a diversity-aware OPD framework. Heterogeneous teacher distillation trains two teachers specializing in popular and unpopular items and routes each training sample to the teacher matching the popularity group of its ground-truth item. The student learns from the routed teacher's probability distribution along its own generated SIDs. Diversity-aware item rollout incorporates ground-truth SIDs of unpopular items into training, covering SID paths that the student rarely generates. Only the student is used for inference. Offline experiments on an industrial dataset and three public datasets further demonstrate its effectiveness, diversity, and robustness. Moreover, we deploy HERO in the "Guess What You Like" module on the 1688 homepage, where a 15-day online A/B test shows a 3.412% increase in gross merchandise value (GMV).
Geyuan Wang, Shunyu Liu, Siyuan Liang, Chenyang Lyu, Dacheng Tao
Large Language Models (LLMs) have shown promise for empowering search agents to tackle complex information-seeking tasks. However, since acquiring external information necessitates web search, these agents are highly susceptible to various adversarial attacks (e.g., malicious information injection or stealthy context manipulation). Existing attack strategies inevitably rely on static and human-crafted templates, failing to emulate the dynamic and complex nature of real-world threats. In this paper, we propose an adversarial self-evolution framework that integrates bi-level optimization between the attacker's strategy evolution and the defender's adaptive mitigation, to expose the vulnerabilities of search agents. Specifically, we introduce the Contrastive Rollout Evolutionary Optimization (CREO) method to drive the directional evolution of the attack strategy population via contrastive evolutionary signals. To provide a stable environment for this continuous evolution, we further construct a controlled sandbox and design fine-grained metrics to quantify internal vulnerabilities. Extensive experiments across various multi-hop QA benchmarks and frontier LLMs demonstrate that the proposed framework yields attack efficacy superior to static baselines, underscoring the urgent need to develop more robust defense frameworks.
Yawen Shao, Jie Xiao, Kai Zhu, Yu Liu, Hongchen Luo, Xueyang Fu, Yang Cao, Wei Zhai 等 9 位作者
Reinforcement learning (RL) has shown strong potential for aligning generative models with human intent, but adapting RL to masked generative models (MGMs) remains largely underexplored. Existing approaches that directly adapt Group Relative Policy Optimization (GRPO) to MGMs leave a foundational conflict unaddressed: MGMs' exploitative decoding nature is inherently at odds with the exploratory diversity that GRPO presupposes. In this work, we reveal a structural decoding bias termed the visual exploration trap: confidence-based sampling trades the exploration of diverse subject realizations for the greedy resolution of low-uncertainty background regions, causing premature collapse of the generative exploration space and starving GRPO of the rollout diversity it depends on. To this end, we propose MaskSense, a novel RL framework for MGMs that confronts the visual exploration trap at both the sampling and optimization stages. Specifically, we introduce a semantic anchored routing sampling mechanism that leverages semantic priors to preserve high-entropy subject token exploration while amplifying intra-group reward variance through exploratory-exploitative routing to yield more discriminative advantage estimates. Furthermore, we design a global entropy transition anchoring strategy to identify the most consequential decoding steps and concentrate gradient updates on them. Extensive experiments on multiple text-to-image benchmarks demonstrate that MaskSense substantially improves upon the base model, with GenEval accuracy improving from 54% to 86% and HPS from 28.89 to 36.39, achieving state-of-the-art performance.
Tianyu Liu, Jun Zhang, Wenjie Huang, Yuhao Shen, Tianqi Xu, Xinyue Hao, Hongyaoxing Gu, Xuyang Liu 等 15 位作者
Speculative decoding (SD) accelerates large language model inference by reducing the sequential cost of token-by-token generation. Advanced SD methods reduce proposal cost with parallel drafters that predict entire blocks in one forward pass, but overall speedup still depends on acceptance length. On-policy distillation (OPD) can push well-trained drafters beyond acceptance-length plateaus, yet conventional pipelines collect supervision through sequential draft-and-verify rollouts. Our key insight is that on-policy supervision need not inherit these dependencies: a draft block can be supervised using only its reference prefix and draft-sampled history. We propose SpecOPD, a branch-parallel OPD framework that parallelizes draft sampling and target supervision across independent branches drawn from the evolving drafter policy at reference anchors. We pack these branches into proposal trees to share target-prefix computation, using tree masks for full attention, branch-consistent indexer groups for sparse attention, and indexed state replay for linear attention to preserve each branch's history. We train the drafter with an E2E-TV surrogate that directly targets acceptance length. Our study spans 8B full-attention (GQA), 27B full-linear (GQA--GDN), and 125B sparse-linear (QSA--GDN) targets, three parallel draft architectures (DFlash, DSpark, and DFlash V2), and four task domains (dialogue, coding, mathematics, and agentic). SpecOPD achieves up to 41.45% higher mean acceptance length, $16.09\times$ supervised-token throughput over sequential-rollout OPD under native configurations, and up to $3.56\times$ generation speedup over autoregressive decoding.
Long-horizon interactions with LLM-based assistants require memory systems that preserve and update user states, preferences, and interaction histories. Existing evaluations report end-to-end QA accuracy and cannot determine whether errors arise from encoding, retrieval, or generation. We introduce EvalMem, an operation-level diagnostic framework with three parallel Examiners. For each query, the Encoding Examiner checks whether the target fact is stored, the Retrieval Examiner assesses whether the native retriever returns usable evidence, and the Generation Examiner tests whether the model can answer from oracle evidence. Their outputs form fine-grained multi-label defect codes. To improve store-level diagnosis, we adapt agentic RAG with a recall-first strategy that searches using both the query and source evidence, increasing recall of present evidence on LoCoMo from 70.2% to 95.6%. Evaluations of seven memory systems on LoCoMo, LongMemEval-S, and dynamic DynaMem-Bench identify retrieval as the most frequently attributed failure layer; in default LoCoMo, retrieval defects reach 22.1%, compared with 7.7% for encoding and 6.5% for generation. Guided by this diagnosis, MemWiki, a search-friendly auxiliary structure built from each system's memory export, improves mean accuracy by 2.5 and 2.3 percentage points on LoCoMo and LongMemEval-S, respectively.