Papers

Digital agents must often adapt to new environments whose interfaces, tools, and failure modes are not fully captured by pretrained models. We introduce RSIAgent, a training-free multi-agent framework for recursive self-improvement through autonomous memory construction. RSIAgent coordinates curriculum, actor, and verifier agents to continually explore the environment, validate outcomes, and retain environment-specific knowledge, including reusable causal relationships between actions, conditions, and consequences. It further adopts a broad-then-deep exploration strategy, combining parallel broad recursive self-exploration for discovering diverse environment structures with focused deep self-exploration for uncovering hard cases, hidden constraints, boundary conditions, and previously unknown causal dependencies. The resulting memory is frozen and can be directly reused for downstream tasks without updating model parameters. Experiments on OSWorld-v2 and Agent's Last Exam show that RSIAgent substantially improves strong open-source models, enabling Kimi-K3 and GLM-5.3 to outperform frontier closed-source models including GPT-6.

Effective communication is a cornerstone of distributed intelligence in Multi-Agent Reinforcement Learning (MARL), yet ensuring that generated messages are both informative and robust to physical constraints remains a significant challenge. This paper introduces Multi-Agent Regularized Communication (MARC), a novel framework inspired by information-theoretic principles of conditional mutual information. MARC employs an attention-based architecture coupled with a unique message regularization mechanism designed to minimize uncertainty regarding future system states, thereby inducing the learning of highly representative communication protocols. Crucially, we evaluate MARC under stringent communication bottlenecks and lossy channels, simulating the real-world constraints of autonomous robotic networks and decentralized systems. Our results demonstrate that MARC significantly outperforms state-of-the-art methods in complex cooperative domains. Furthermore, we provide a deep analysis of message characteristics, proving that MARC maintains high operational performance even under significant data compression, offering a scalable path for deploying intelligent agents in resource-constrained environments.

The surge of large language model (LLM) applications on personal devices imposes massive, bursty workloads on cloud serving infrastructure. While prefill-decode disaggregation improves throughput and scalability, memory-bound decode instances often suffer from persistent load imbalance, as output lengths are unknown when requests arrive at the cloud. To address this, we propose MAPS, a Memory-Aware Predictive Scheduling framework tailored for disaggregated LLM serving. MAPS performs device-assisted speculative output length prediction overlapped with cloud-side prefilling, incurring negligible latency overhead. To handle generation uncertainty, MAPS applies uncertainty-aware calibration to derive output-length upper bounds with target coverage, enabling safe scheduling decisions. Building on these bounds, MAPS employs a hierarchical global-local scheduling strategy to mitigate inter-decoder queue buildup and intra-decoder head-of-line blocking. Extensive experiments on two real-world workloads and two LLMs show that MAPS significantly outperforms three state-of-the-art systems, reducing average end-to-end latency by 42.6 and tail latency by up to 84.8.

Accurate cell detection and classification are crucial for pathological analysis, directly affecting diagnostic accuracy and treatment planning. To capture complex cellular interactions beyond visual appearance within the tumor microenvironment, several approaches have employed graph neural networks to model spatial and relational patterns among cell nuclei, yielding promising results. However, these methods typically adopt a two-stage paradigm of visual extraction followed by relational modeling, which necessitates separate tuning for each stage, thereby increasing pipeline complexity and hindering end-to-end joint optimization. In this paper, we propose an end-to-end framework for cell detection and classification that jointly models patch-level visual representations and instance-level interactions, which incorporates a dynamic graph construction module and an instance-aware graph network. Specifically, the graph construction module dynamically builds the graph structure using learnable queries derived from patch-level features as cell instance representations, with adjacency defined by integrating feature similarity and spatial distances. The instance-aware graph network performs adaptive instance filtering and feature reorganization, aggregating them over the cell graph into a topological latent state for a selective state-space transition driven by visual cues, fusing appearance and relational evidence. When evaluated on multiple datasets with different staining protocols for cell and nucleus detection, our method significantly outperforms existing approaches in both detection and classification performance. The code will be released at https://github.com/RuochenLiu23/IGM.

We study the adaptation of pretrained language models to univariate time-series forecasting through a parameter-efficient transfer learning framework, with the goal of understanding which design choices drive effective cross-modal transfer. While language models operate on discrete textual tokens, time series consist of continuous numerical observations with temporal dependencies. To bridge this modality gap, we project fixed-length time-series patches directly into the embedding space of a pretrained GPT-2 backbone, bypassing textual tokenization and treating the Transformer as a generic sequence encoder. Through controlled ablation studies on seven benchmark datasets spanning energy, weather, traffic, and finance, we analyze the effects of (i)~representation strategy (continuous embeddings versus textual serialisation), (ii)~adaptation regime (frozen backbone versus partial or full fine-tuning), (iii)~architectural components such as adapters, pooling strategies, and prediction heads, and (iv)~input context length. Continuous patch-based embeddings consistently outperform textual prompting and randomly initialised backbones. The adapted pipeline attains MASE within the range of specialised forecasting architectures while updating less than 1% of total model parameters. Results further indicate that freezing the pretrained backbone and training lightweight projection and adapter modules provides a favourable accuracy–efficiency trade-off with stable behaviour across varying context lengths.

Large Language Models (LLMs) primarily perform inference at the token level, resulting in substantial memory overhead and compromised computational efficiency. In this paper, we propose a Dynamic Semantic Extraction and Inference (DSEI) framework, which achieves segment-level inference within the latent space through a two-stage training strategy. First, we construct a Dynamic Semantic Autoencoder (DSAE) via self-supervised learning. DSAE dynamically extracts segment-level semantics and compresses them into compact latent representations via adaptive semantic weighting and gated fusion. Subsequently, we integrate the DSAE into the LLM architecture and train the model to infer over dense latent space. DSEI substantially reduces both input and generation sequences and significantly enhances inference efficiency. Extensive experiments conducted on the Wanjuan dataset demonstrate that DSEI reduces perplexity by 48% compared to static sentence-level latent inference baseline. Furthermore, compared to standard LLMs using token-level inference, DSEI accelerates inference speed by 2.5$\times$ and reduces memory overhead by 90%.

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.

Pretrained driving vision-language models (VLMs) integrate visual, route, language, and driving context into rich driving priors, yet their representation objectives remain separated from continuous driving planning. Existing methods typically begin trajectory generation only after the VLM has formed a final condition, leaving depth-wise condition computation outside the stepwise formation of trajectory state. We introduce DiffAdapterVLA, which realizes Planning in the Backbone: it injects explicit trajectory tokens into selected VLM late layers, bringing trajectory state into backbone forward computation, where it co-evolves with driving conditions at different depths. Lightweight layer-wise DiffAdapters organize this computation into recursive trajectory refinement, while asymmetric joint attention preserves directed guidance from the condition stream to trajectory planning. By placing planning within existing backbone computation rather than relying on an independent trajectory planner, DiffAdapterVLA adapts only lightweight trajectory modules to turn existing driving priors into efficient continuous planning capability. NAVSIM results show that it achieves high-quality closed-loop planning with low end-to-end latency using few trainable parameters, and demonstrate that jointly evolving trajectory state and depth-wise driving conditions in VLM late-layer computation effectively realizes continuous trajectory planning.

Frontier models score well on shallow document/chart reading tasks. In a controlled data-room audit, moving evidence into buried conditions reduced accuracy, increased forced declarations, increased tool calls, and increased cost per correct answer. Confidence and benchmark calibration did not fully capture wrong answers; a documented production incident shows fabricated structural claims can be mixed with accurate numeric tables. Agentic evaluations need claim-level receipts (statement-level provenance, not answer-level scores), condition-aware scoring, and human-adversarial verification - an auditing discipline, not a leaderboard. The setting we measure is financial due diligence; the setting we are building toward next is defense staff work, where the same buried-evidence shape appears. In both, the model is not a party to the consequences; the person who signs is. In plain terms: in the documented cases we examine, agents can pair accurate numbers with confident fabricated explanations, and the burden of proof must therefore move from the model to the evidence trail.

Safe reinforcement learning (RL) is commonly formalized as a Constrained Markov Decision Process (CMDP), in which an agent maximizes expected reward while keeping its expected cumulative cost below a specified safety bound. Existing safe RL benchmarks predominantly report whether an algorithm is safe on average, following this expectation-based guarantee. We argue that this convention is insufficient to reliably characterize an algorithm's true safety: it fails to capture how often and how severely the safety bound is violated, whether this holds consistently across tasks and safety bounds, and whether training-time behavior is representative of behavior of the final converged policy. Therefore, we introduce (i) evaluation metrics for safe RL that address each of these concerns and in addition allow for aggregation across tasks and safety bounds. We furthermore define (ii) a safety tier system to systematically categorize and compare algorithms in terms of safety and reliability at both training and for a final policy. Using this framework, we provide (iii) an empirical safety evaluation across multiple safety navigation tasks. Our results show that aggregate metrics, distributional reporting, and task- and safety bound-specific results each reveal information the other metrics cannot. We therefore recommend reporting all three jointly, rather than compressing this information into a single value, as is common practice. We provide SafeRLEval, an open-source evaluation suite to support the reliable characterization of safety in future safe RL research.

We introduce the Universe of Universes (UoU) framework, which treats the full ecosystem of major large language models (LLMs) as a structured retrieval corpus and proposes a compositional Automated Reasoning (AR) and Machine Learning (ML) architecture for cross-model retrieval-augmented generation. The central contribution is the formal characterization of the Benefit Yield Function (BYF), the marginal performance gain per additional model added to an ensemble, and the identification of the implosion threshold {\theta}*: the ensemble size at which BYF crosses zero and aggregate performance begins to degrade. Existing LLM ensemble and mixture-of-agents systems treat models as responders and aggregate outputs, but do not study performance as a function of ensemble size N across the full model universe. Benchmark research confirms performance plateaus at the individual model level; model collapse literature establishes that iterative training on AI-generated outputs degrades individual model distributions. Neither body of work formalizes the ensemble-level implosion threshold, models Epistemic Hereditary Drift (EHD) at the ecosystem level, or treats AI manufacturing velocity as a co-variable of {\theta}*. The framework has direct implications for DoD multi-model AI acquisition policy and the emerging science of testing AI-enabled systems.

Urban region representation learning commonly combines heterogeneous data sources, such as mobility flows, points of interest, and land-use information, to support tasks including mobility analysis, public safety forecasting, and service demand estimation. Existing multi-view methods typically improve region embeddings by strengthening interactions across views. However, such methods often overlook view-specific regional structures and may propagate correlations induced by shared latent factors, which can reduce the stability of downstream predictions. To overcome this major limitation, we propose CURE, a confounder-aware framework for multi-view urban region representation learning. CURE first encodes each view with its regional graph structure, estimates a shared latent component, and then reduces its projected influence before cross-view interaction. A hierarchical graph-aware fusion module subsequently aggregates the residual view representations using local and global regional contexts Experiments on three real-world cities show that CURE improves predictive performance, remains robust under missing and noisy input views, and provides reliable cross-view integration through shared component separation and context-dependent view weighting.

Universal multimodal embedding (UME) maps multimodal inputs into a shared embedding space for diverse retrieval tasks. Recent methods improve embeddings through Chain-of-Thought (CoT) reasoning optimized with GRPO using retrieval rewards. However, existing methods overlook the mismatch bettween candidate-aware retrieval supervision and input-only CoT generation: (1)trajectory-level rewards convey retrieval outcomes without explicitly identifying the input-supported evidence that distinguishes the positive from hard negatives; (2) input-only generation cannot directly assess whether further reasoning improves retrieval, potentially producing redundant CoTs with substantial latency. To bridge this gap, we propose Reason What Matters (ReWAM), a retrieval-grounded framework that aligns candidate-aware supervision with input-only generation. Specifically, we introduce Retrieval-Aware Self-Distillation (RASD), which extracts privileged guidance from input-supported facts and evidence distinguishing the positive from hard negatives. Conditioned on this guidance, an on-policy self-teacher provides token-level feedback to refine credit assignment, directing policy updates toward retrieval-relevant reasoning grounded in the input. We further propose Retrieval-Adaptive Inference (RAI), which learns a retrieval-aware stopping criterion from prefix-level retrieval feedback. It stops redundant reasoning without candidate access and uses speculative decoding to further reduce CoT latency. Extensive experiments on MMEB-V2 and MRMR demonstrate that ReWAM achieves state-of-the-art retrieval performance while delivering up to 5x the inference throughput of competitive explicit-CoT UME methods. ReWAM thus enables high-quality retrieval through efficient input-only reasoning, making explicit CoT practical for corpus-scale multimodal retrieval. The code will be publicly available.

Binding the audit flag in Reflexion-style agents — without changing the auditor — reduces attack success rate substantially, reaching near zero on models whose flags parse cleanly. This single control-flow change exposes the enforcement gap: the controller receives a safety flag and executes anyway. Separating detection probability $p_d$ from enforcement probability $p_e$ establishes that $p_e \approx 0$ by default across every framework we tested, making detection quality formally irrelevant to security when enforcement is absent — a finding consistent with the spontaneous collapses recorded in unsupervised frontier-agent deployments~\citep{emergence2026}. Residual attack success concentrates where flags are unparseable or auditors leak; an RL-trained enforcement controller handles hedged and malformed verdicts that rule-based parsing cannot, cutting ambiguous-critique failure to a fraction of the rule-based baseline. Concurrent filtering and information-flow defenses address detection, not enforcement, leaving the binding constraint untouched. The Audit Enforcement Specification (AES) packages three concrete requirements that close each residue independently; each primitive is adoptable without redesigning the host framework, and no deployed framework currently implements any of them.

Automatic Item Generation (AIG) is pivotal for personalized education, yet guaranteeing the pedagogical value of generated items remains a bottleneck. Existing Item Quality Assessment (IQA) methods typically rely on unscalable manual reviews or shallow stem-based metrics, failing to capture the reasoning process required for mathematical problem-solving. To bridge this gap, this paper proposes Process-based Item Quality Assessment (ProIQA), a process-aware framework for fine-grained quality assessment of math items. We first formulate IQA across three heterogeneous dimensions, including knowledge concepts, difficulty, and disciplinary competencies, under a unified process-aware perspective. Based on this formulation, we construct a process-enhanced IQA resource by augmenting original item data with structured reasoning trees derived from raw solutions. Technically, ProIQA leverages Large Language Modelsto construct hierarchical reasoning trees and employs Graph Neural Networks (GNN) to encode their topological dependencies and procedural semantics. The resulting solving representation is fused with stem semantics through a dual-view (“Stem + Solving”) architecture, enabling comprehensive assessment across learning objectives. Extensive experiments on K12 mathematical datasets show that ProIQA effectively captures process-oriented features, offering a scalable data-driven solution for evaluating AIG outputs in intelligent education systems.

Artificial intelligence threatens to outrun human understanding and control. New mathematics is needed to design AI that is legible, steerable, and cooperative with humanity. I organize this invitation by mathematical field, so you can turn straight to your own: logic and game theory for cooperation; probability for agency and world-models; algebra and representation theory for learned features; analysis and geometry for generalization and training dynamics. Each section ends with an open problem that is accessible to a working mathematician with no prior experience in AI safety.

Environmental, Social, and Governance (ESG) reporting is critical for corporate accountability, with Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG) offering strong potential to automate KPI extraction. However, open-source LLM performance in domain-specific ESG tasks remains insufficiently understood. This paper evaluates open-source LLMs in ESG contexts using a structured framework and evaluation resource based on 498 real-world ESG reports from EU-listed companies (2010-2024). We evaluate seven open-source models (2B to 30B parameters) – glm-4.7-flash, nemotron-3-nano:4b, qwen3:4b-instruct, gemma3:4b, gemma4:e4b, gemma4:e2b, and ministral-3:8b – using 100 persona-based synthetic QA pairs covering ESG information needs. System performance is assessed via RAGAS metrics, including contextual recall, precision, relevance, faithfulness, answer relevancy, and factual correctness. Results show notable performance variations across architectures. Retrieval performance is strong across models (context recall around 0.58-0.61, context precision around 0.78-0.81, context relevance 0.965-0.985). Generation diverges most on faithfulness (0.607-0.822) and least on answer relevancy (0.760-0.881): glm-4.7-flash leads in faithfulness (0.822), qwen3 in factual correctness (0.449), and ministral-3 in answer relevancy (0.881). Low overall factual correctness (0.387-0.449) highlights the need for domain-specific fine-tuning. This work provides data-driven guidance for deploying open-source models in ESG reporting.

Discrete diffusion models are a prominent family for graph generation, but standard class-conditional mechanisms embed the class signal in the denoiser during training, tying the conditioning mechanism to the trained model. Classifier guidance avoids this coupling in continuous domains by steering a frozen model with a classifier's gradient, but discrete graph diffusion samples discrete edge states, so gradients cannot propagate through the sampled graph. We introduce ProtoGuide, a post-hoc, backbone-agnostic framework that recovers an analogous mechanism. At each reverse step the denoiser's per-edge output is relaxed into a differentiable soft adjacency, embedded by a frozen Siamese graph neural network, and scored against a target-class prototype and its nearest competitor; the resulting per-edge gradient, damped by a cosine schedule, is injected back into the denoiser output. All components stay frozen, so guidance is retargeted by supplying a different prototype. On five classes of real-world networks and two architecturally different backbones, EDGE and DiGress, ProtoGuide raises macro classification accuracy from 50.7% to 73.5% and from 73.6% to 83.8%, and outperforms DiGress's built-in conditional training under our configuration. Gains are largest where the unguided models are weakest, and are not uniform across classes. Per-graph coverage remains high in most settings, while distributional effects are class-dependent. A Best-of-N selection baseline matches this accuracy given enough oversampling, but at a substantial cost in graph diversity. An independently initialized classifier, a directionality test, and a few-shot analysis support target-directed steering and robustness to very small support sets.

Recently, Large Language Models (LLMs) have gained significant attention due to their strong language understanding and generation capabilities, demonstrating impressive reasoning abilities as well as effective utilization of external knowledge. Many studies have proposed methods that specialize in improving performance for individual tasks. However, ironically, only a limited number of attempts have explored general-purpose, task-agnostic methods. In this work, we present a unified framework integrating reasoning and Retrieval-Augmented Generation (RAG) tasks. We further propose Controllable White-Box Meta-Prompting (CWM), a low-cost white-box method for adaptive RAG tasks previously dominated by black-box approaches, without requiring external decision modules or multi-sampling. CWM achieves state-of-the-art performance on three adaptive RAG benchmarks across recent LLMs, including GPT-oss-20b, Qwen3-14b, and Llama3.1-8b, while also demonstrating strong generality by extending to reasoning tasks. In addition, CWM provides controllability by enabling retrieval decisions to be regulated through the manipulation of internal model signals. Our code is available at https://github.com/JeongEunhye00/CWM.

We propose a pre-fine-tuning probing method for Parameter-Efficient Fine-Tuning (PEFT) layer selection, aiming to obtain more stable and higher gains with fewer trainable parameters when adapting large vision–language models (VLMs). Unlike the common practice of applying LoRA and other adapters to all layers at once—where layer selection often relies on heuristic rules—we focus on the vision encoder and directly evaluate the "adaptability” of each Transformer layer. Specifically, we characterize each layer from two perspectives: (i) the statistical properties of its Q/K/V projection weights (e.g., norms and condition numbers); (ii) robustness under controlled parameter perturbations. We then systematically compare these indicators with the downstream performance gains brought by applying PEFT to a single layer. Across experiments covering seven benchmarks and five PEFT variants, we observe a consistent correlation: layers (or matrices) with larger weight norms and higher condition numbers are usually more robust to perturbations and are more likely to yield larger fine-tuning gains. These results show that distribution-statistics analysis and perturbation tests before fine-tuning can provide practical signals for adaptation-layer selection, thereby maintaining or improving performance while reducing trainable parameters.

Founders face two linked decisions: whether to pursue an idea before founding, and which operating actions and capital partners fit afterward. We present a public-data decision-support toolchain combining time-bounded proposal profiling, market and moat checks, and deterministic aggregation with auditable investor-company event chains for retrospective analysis. Pre-founding: (a) After threshold selection on 198 development companies, the frozen pipeline achieves F0.5=0.5357 [0.412, 0.655] on an independent, row-disjoint 198-company validation sample. On the combined 396 rows, the Full Pipeline scores 0.6301 versus 0.2734 for a paired Raw LLM baseline. Post-stratification of 1,027 completed cases in a separate scale cohort yields 0.6506 [0.598, 0.707]; the run remains incomplete. A 377-row composition-matched check yields 0.6573. (b) The AI-inference study identifies distribution-layer businesses as a replicable path to independent profitability with a limited revenue ceiling, and frontier-model ownership as a path to capital-market upside at exceptional capital cost. Post-founding: (a) Public sources support auditable event-chain analysis. (b) In the chip-company study, sustained product, customer, and supply-chain progress is associated with better observed outcomes; financing alone does not establish operating progress. (c) Financing comprises 79% of confirmed visible post-investment actions. Evidence tentatively favors acquisition-experienced strategic corporate investors for acquisition-oriented founders and financing-led institutional VCs with fewer observed control events for independence-oriented founders. Findings are developmental and observational, not causal guarantees or investment advice. We release shared ontology, provenance-bearing EventChain data, schemas, benchmarks, and executable skills for audit, reuse, and extension.

Large Audio-Language Models (LALMs) have substantially advanced general audio understanding, yet they remain limited in fine-grained temporal perception, particularly in precise event localization. Existing approaches primarily post-train LALMs to predict event boundaries as timestamp tokens. However, this generative formulation lacks explicit correspondence between the timestamp predictions and fine-grained acoustic evidence, limiting the precision and reliability of temporal localization. To address this issue, we augment the LALM with a dedicated frame-level grounding model while leveraging its semantic modeling capability to represent the event query. Specifically, the frozen LALM encodes the event query with audio as context, and the grounding model combines these query representations with fine-grained audio features to localize the target event at the frame level. Extensive experiments across diverse temporal grounding benchmarks demonstrate strong and consistent improvements over existing methods. Further evaluation shows that the grounding model can provide temporal evidence to support downstream reasoning.

Automatic prompt optimization (APO) improves language-model programs by revising prompts from task feedback, yet it typically holds its training data fixed. Repeatedly optimizing against the same instances confines feedback to weaknesses already represented in those data, leaving related failure conditions unexplored. We therefore view each failure as a dual signal: it indicates both how the prompt should be revised and what new training evidence should be synthesized. We introduce FORGE, a failure-guided framework that co-evolves prompts and training data. FORGE abstracts imperfect executions into reusable failure modes and synthesizes new training data through four complementary mutation strategies. Verified instances are fed back into prompt search, allowing updated prompts to expose the next data needs. Across eight heterogeneous benchmarks, FORGE improves the aggregate score over the unoptimized baseline by 16.52 percentage points and outperforms all evaluated APO baselines. The synthesized data also transfer beyond FORGE: in a transfer study, they improve all nine APO comparisons by 2–9 points and all three GRPO comparisons by 4–8 points under matched optimization budgets. These results establish failures as a shared interface between prompt optimization and data synthesis, and show the benefit of jointly adapting what a model is instructed to do and what it learns from.

Generative AI writing assistants and the Large Language Models (LLMs) that power them are increasingly part of how voters gather information before elections. With growing evidence that they influence users' opinions, it is increasingly important to understand the views and positions of these tools. To better understand these views, we examine the stances supplied by six LLMs on a variety of Swedish-language writing tasks ahead of the 2026 Swedish parliamentary election. We cross 107 policy propositions with 77 writing templates and neutral, positive, and negative prompt framings, producing 24,717 prompts per model and 148,302 responses. To study these, we look at the models' default stance tendencies, compare how they respond to similar issues, and compare their responses with those of each of Sweden's eight parliamentary parties on the same issue. We find that Claude, DeepSeek, Gemini, and Mistral have similar profiles; ChatGPT more often supplies neutral or ambivalent text; and Grok differs most on topics such as migration, crime, and gender. When comparing the political parties, we find that the Social Democrats are closest to all six models. Still, after correcting for multiple comparisons, none of the within-model differences in party distances remains significant. Overall, we find that no model has a clear preference, nor a clear preference for a party, but that this depends on the specific issue or task the user asks about.

Table extraction from texts is an important task for information systems, and recent approaches that prompt large language models (LLMs) with instructions have drawn great attention for their strong performance. Existing works have assumed the input texts to be table descriptions or specialized documents. However, these efforts have largely overlooked another prevalent category of texts, commonly found in news reports and social media: naturally occurring texts. Extracting tabular information from such texts poses two distinct challenges. First, high variability and the absence of explicit structural cues make fixed heuristic LLM prompts limited in precisely delineating extraction boundaries. Second, manually predefined schemas cannot capture open-ended, unseen attributes in naturally occurring text. In this paper, we propose a framework, TEAR, to address these challenges. It comprises two synergistic workflows: a Table Extraction Workflow that dynamically adapts instructions to overcome the limitation of heuristic instructions, and an Attribute Recommendation Workflow that discovers new attributes from texts to complement the heuristic schema. To our knowledge, TEAR is the first framework that supports automated text-driven attribute recommendation, enabling exploratory schema design for table extraction. To evaluate TEAR, we establish the benchmark for table extraction and attribute recommendation on naturally occurring texts, including two real-world datasets, manual annotations, appropriate metrics, and baseline comparisons. Experiments show that TEAR achieves state-of-the-art performance on both tasks, and the recommended attributes effectively enhance extraction performance in exploratory scenarios.

Speaker recognition neural networks recognise speaker identities from input utterances by learning latent representations (i.e. speaker embeddings). However, these networks' internal mechanisms remain largely opaque, motivating research in explainable artificial intelligence (XAI) to understand them. Existing studies have analysed how speaker embeddings are organised, but rarely frame these analyses within XAI. This work proposes to explain and interpret the organisation of speaker embeddings from an XAI perspective. To this end, we apply a hierarchical clustering algorithm, Single-Linkage Clustering (SLINK), to analyse whether our prepared speaker embeddings naturally form clusters with hierarchical relationships. The resulting hierarchical organisation (i.e. hierarchical clusters) is evaluated using the Cluster-Class Matching (CCM) method. Moreover, we propose a new method, termed Hierarchical Cluster-Class Matching (HCCM), to identify which hierarchical clusters best match individual semantic classes like male and conjunctive semantic classes like UK & male, thereby interpreting the clusters using their matched classes. We quantify the matching degree with a new metric called the L-score, which makes imperfect matches diagnosable. HCCM's results reveal that the hierarchical clusters analysed by SLINK are well interpreted using classes related to speaker identity, gender, and nationality, showing the semantics inside the hierarchical organisation of the examined speaker embeddings.

Real-world visualization requests are routinely ambiguous, incomplete, or factually incorrect, yet existing Text-to-Visualization (Text-to-Vis) systems assume well-specified inputs and produce charts in a single pass. When queries are imperfect, a system must interact with the user to recover the true intent, but no benchmark or method supports this dynamic process. We introduce VisInteract, a new paradigm that reframes Text-to-Vis as interaction-driven intent recovery, and VisInteract-Bench, to our knowledge, that is the first benchmark for dynamic interactive Text-to-Vis, featuring controlled imperfection injection, a leakage-controlled User Agent for realistic multi-turn feedback, and dual-perspective (code and chart) automated evaluation. On the algorithmic side, we propose Vis-MCTS, a Monte Carlo Tree Search (MCTS) enhanced method, introducing improvements over classical MCTS, that Progressive Widening to tame the unbounded tool-argument space in tree search, cross-rollout information sharing so clarifications and critiques benefit the entire search tree, and Dimension-Aware Reward Decomposition that routes scalar user feedback along data-fidelity, visual-design, and intent-alignment dimensions to resolve credit assignment across heterogeneous actions. Extensive Experiments across two LLM backbones show that Vis-MCTS consistently outperforms all Text-to-Vis baselines, improving end-to-end task success by $13.40\%$–$16.27\%$ over the strongest interactive baseline and by more than $5\times$ over non-interactive ones.

Spatio-temporal traffic forecasting is a fundamental big data analytics task for intelligent transportation systems, where massive urban sensor streams exhibit heterogeneous, non-stationary, and structurally dynamic patterns. Although recent deep learning and large language model (LLM)-based methods have advanced traffic forecasting, they often remain temporally centered and lack effective coordination of temporal, spectral, pairwise spatial, and higher-order structural cues under evolving traffic regimes. To address this heterogeneous dependency coordination problem, we propose STHMoE, a Spatio-Temporal Hypergraph-Enhanced Mixture of Experts framework for urban traffic data forecasting. STHMoE decouples traffic dynamics into frequency-domain, time-domain, spatio-domain, and higher-order spatial representations, which are modeled by prompt-guided heterogeneous experts built upon a partially frozen LLM backbone. The first three experts leverage domain-specific statistical prompts, while the higher-order spatio expert uses a structural placeholder prompt and obtains dependency information from an adaptive hypergraph module. To capture evolving spatial structures in traffic data,, STHMoE jointly learns first-order graph dependencies and higher-order group interactions without predefined topologies. An entropy-aware MoE router with coefficient-of-variation load balancing adaptively fuses expert outputs while improving expert utilization and routing confidence. Experiments on 10 real-world traffic benchmarks show that STHMoE achieves competitive performance against temporal, spatio-temporal graph, and LLM-based baselines.

Large language model (LLM) driven multi-agent systems have shown promise in complex clinical reasoning, yet existing approaches rely on static strategies and lack persistent clinical memory, preventing self-evolving from prior diagnostic successes and failures. We present EMR, a self-evolving medical multi-agent system via Experience Mining and Reuse. EMR introduces a hierarchical clinical experience library that organizes accumulated knowledge into three levels: clinical principles, diagnostic patterns, and representative cases. During inference, EMR emulates multidisciplinary consultation: a planner agent coordinates domain-specific department agents for specialized reasoning, while a summary agent synthesizes their analyses into a final decision. Critically, EMR automatically extracts correct diagnostic insights and failure-related warnings from multi-agent reasoning trajectories, incrementally updating the experience library to guide future cases. Experiments on medical reasoning benchmarks demonstrate that EMR consistently outperforms state-of-the-art medical multi-agent baselines. Further analysis reveals that the hierarchical experience enables cross-specialty generalization and transfer across diverse LLM backbones, offering a scalable and in

Looped Transformers have recently demonstrated strong performance in both reasoning and language tasks by reusing a shared set of parameters across multiple iterations, achieving parameter efficiency without sacrificing representational power. Besides, looped Transformers perform inference directly in the latent space (latent reasoning) to reduce the number of tokens consumed during inference, thereby achieving improved sample efficiency. However, these models typically apply a fixed recursion depth uniformly to every token, leading to suboptimal compute allocation and leaving significant efficiency gains on the table. In this work, we propose token-level elastic-depth looped transformers: T-LoopFormer. T-LoopFormer uses dynamic token-choice routing, which enables each token to adaptively determine its own number of loop iterations based on its hidden state, which could improve the token generation accuracy. Moreover, we further introduce recursion-wise KV cache, which maintains an independent key-value cache for each recursion loop of active tokens; this design ensures that tokens at different depths only attend to their corresponding cached states, effectively enabling faster autoregressive decoding. Extensive experiments show that T-LoopFormer achieves robust performance on a wide range of language modeling and zero-shot reasoning tasks, and our model could reach the lowest decoding latency, which validates the effectiveness of token-choice router and recursion-wise KV cache. Code is available at https://github.com/YuMingQian1234/T-LoopFormer