Papers

LLM-based query expansion increasingly conditions reformulation on documents retrieved from the target corpus, yet most work focuses on how to generate expansions rather than which documents the model should read. We propose EviQE, which aggregates documents retrieved by multiple reformulators, selects a compact evidence set, and uses it for one grounded expansion step. This separates evidence selection from generation and treats reformulators as complementary retrieval perspectives. Across three TREC DL and five BEIR benchmarks, reformulators frequently retrieve distinct relevant documents, so pooled candidates provide higher relevant-document coverage than any individual source. The strongest gains come from relevance-based evidence selection: LLM-Score consistently outperforms direct reformulation, cold-start expansion, and single-source seeded expansion. Additional retrieval-generation rounds provide little benefit once strong conditioning evidence has been selected and can reduce effectiveness.

Medical volume visualization requires selecting regions of interest (ROIs) and carefully controlling their relative visual emphasis according to a given clinical intent. Implementing these decisions in conventional workflows demands substantial clinical and visualization expertise and often involves trial-and-error optimization. Recent agentic systems have introduced natural-language interaction and autonomous visualization operations but largely rely on MLLM-based inference throughout the workflow. Although MLLMs encode broad medical knowledge and provide strong reasoning capabilities, such inference may be suboptimal for medical volume visualization, potentially leading to clinically incomplete interpretations of user requests and unreliable ROI identification and visualization optimization. In this work, we present MedVA, an end-to-end neuro-symbolic agentic system for medical volume visualization that addresses these limitations through three complementary agents. The neuro-symbolic intent formulation agent refines MLLM-based interpretations of natural-language requests through symbolic reasoning over established clinical knowledge, which provides more complete, clinically grounded ROI specifications than MLLM-only reasoning. The multi-model ROI identification agent directly identifies semantically specified ROIs in the original volume by leveraging complementary large-scale pretrained medical segmentation models. The objective-driven visualization optimization agent explicitly evaluates ROI visibility and occlusion in the original volume using a volume-based visibility objective. Extensive agent-level and system-level evaluations across diverse medical datasets and interaction scenarios support the effectiveness of the individual agents. A formative user study further indicates high usability and practical value among users with different levels of expertise.

We study weighted edit distance between two strings of total length $n$, where edit costs are induced by an arbitrary metric. For equal-length inputs, Kuszmaul (2019) gave an $O(n^\delta)$-approximation with $\widetilde{O}(n^{2-\delta})$ running time for every fixed $0 < \delta < 1$. We give the first constant-factor approximation for weighted edit distance over arbitrary metrics in strongly subquadratic running time. For every $0 < \varepsilon \le 1$, our randomized algorithm runs in $\widetilde{O}(n^{7/4}/\varepsilon^8)$ time and returns a $(3+\varepsilon)$-approximation with probability at least $1-n^{-10}$. The algorithm allows unequal input lengths and places no bound on the ratio between edit costs.

Agentic serving can consume orders of magnitude more tokens than chatbot workloads, stressing both KV-cache capacity and decode-time bandwidth. Most KV-eviction methods score cached keys against representative queries drawn from the most recent tokens, assuming future attention resembles recent attention. We show that agentic generation violates this assumption: future queries form a mixture over think, act, tool, and others phases, and principal-angle analysis shows these components occupy measurably different query subspaces, so recency representatives systematically undervalue keys that upcoming phases will need. We propose AGENTKV, which maintains a small query buffer per phase and scores cached keys against their union. We further implement AGENTKV in a persistent multi-turn serving path that carries compressed KV state across turns and compacts retained KV pages online. Across two models, six task domains, and three KV budgets each, AGENTKV improves task score by 5.5 points on average over R-KV and 5.3 over Tri-attention. Relative to upstream full-KV SGLang, AGENTKV improves output-token throughput by up to 1.80x. Code: https://github.com/LiuTaowen-Tony/agentkv.

Counting equivalent pairs is a common way to report transformation-audit coverage, but it can substantially overstate the constraints imposed by an audit: pairs generated from the same semantic object are correlated, and complete orbit graphs contain algebraically redundant edges. We therefore distinguish four complementary quantities—edge count $m$, effective contrast rank $s$, population support rank $r$, and graph spectral gap $\eta$—and characterize their roles in audit coverage and deployment reliability. Under a rank-$r$ Gaussian contrast model, a population-invariant calibrated reader exists exactly when the anchor has a component in $\ker T$. When an audit has rank $s < r$, its unobserved risk is $R^\star/U$, with $U \sim \operatorname{Beta}((r-s)/2,s/2)$; when $s \ge r$, exact calibrated interpolation is infeasible. The same distinction appears in orbit topology: a spanning tree imposes the same exact-null constraints as a complete graph, while a sharp graph Poincare inequality propagates edge-level drift to an entire orbit at a cost proportional to $1/\eta$. Cyclic audits can additionally yield zero pair-level leave-one-out error without holding out any semantic object. To address these failures, we derive exact block-Woodbury leave-one-orbit-out updates and introduce a source-disjoint deployment gate over finitely many candidate readers. The gate retains the original reader unless uncertainty bounds certify lower drift within a prescribed clean-utility budget. etc..

The multimode fiber is among the thinnest imaging conduits available, carrying hundreds to thousands of spatial modes through a cross-section comparable to a human hair, but its endoscopic capabilities are currently limited by the sensitivity of the transmission matrix to the fiber's deformed state. Proximal-only recovery of the fiber's transmission matrix is an appealing approach for enabling general use multimode fiber endoscopy, and within the last decade, machine learning techniques have been applied to both single-ended and double-ended transmission matrix recovery tasks. We present a new approach to this interdisciplinary problem and show that neural networks can generalize to recover transmission matrices of an arbitrarily deformed graded-index multimode fiber from proximal measurements alone.

We present a primitive-informed sampling-based model predictive control (MPC) framework for multi-fingered dexterous manipulation. Sampling-based MPC evaluates candidate control trajectories through forward simulation without requiring gradients through complex contact dynamics. However, directly sampling these trajectories in the high-dimensional joint space of a dexterous hand is inefficient and makes performance strongly dependent on the sampling distribution. Our framework biases sampling using low-dimensional manipulation primitives that encode coordinated finger motions, while simultaneously optimizing joint-level residuals to adapt these motions to the current hand-object configuration. Task-related rollout constraints reject infeasible trajectories during forward simulation, improving the effective use of the sampling budget. We evaluate the approach on a 16 DoF Allegro hand using a synchronized MuJoCo digital twin. Ablations show that both the primitive and residual are necessary for reliable continuous in-hand rotation, that increasing the sampling budget alone does not recover this coordination, and that rollout constraints substantially improve success rate. A primitive extracted for one object size transfers to other sizes and remains effective under model mismatch. The framework further supports grasping, object reorientation, and coordinated arm-hand manipulation, using primitives extracted from both a simulation-trained policy and human hand-motion data.

We study the minimum coefficient degree $N_{k,\F}(G)$ of a Nullstellensatz certificate for Bayer's $k$-coloring equations, where the characteristic of $\F$ does not divide $k$. If $J$ is a \HJ\ join of non-$k$-colorable graphs $G,H$ and $m=\max\{N_{k,\F}(G),N_{k,\F}(H)\}$, then $N_{k,\F}(J)\leq m+k$. When deletion of the selected edge makes each input $k$-colorable, we also have $N_{k,\F}(J)\geq m$; the degree congruence then gives $N_{k,\F}(J)\in\{m,m+k\}$. This partially answers a question of Li, Lowenstein, and Omar. For three-coloring over $\F_2$, we construct an infinite $4$-critical family of exact degree seven, attaining the bound at input degree four. In contrast, every graph constructed from $K_4$ solely by \HJ\ joins has degree $O(\log n)$ and a certificate with polynomially many terms: joins preserve treewidth at most three, and balanced separators yield low-degree certificates. Additional vertex identifications are excluded from this obstruction. We classify all single identifications of the $25$-vertex base graph; exactly $36$ preserve degree seven, producing $24$-vertex $4$-critical graphs of treewidth four. A compressed self-join at adjacent true twins prevents degree loss and gives a repeatable rule adding four vertices per round. The rule does not establish degree amplification or preservation of criticality. Exact witnesses and standalone verification programs accompany the finite results.

We predict single-sequence model throughput from GGUF metadata using roofline-shaped predictors with quantization-specific scale factors fitted on reference models. The scored cohort comprises 318 phase-depth measurements from 53 host-file configurations on two Apple M4 Max systems and an NVIDIA RTX 5080. On host-specific held-out sets of four, five, and two configurations, an active-parameter decode model obtains 13.1%, 14.4%, and 36.1% mean absolute percentage error (MAPE), versus 49.4%, 55.3%, and 51.9% when charging total parameters. Leave-one-host-out coefficients fitted on the other two systems yield 11.6%, 16.8%, and 36.0% test MAPE. A low-bit model ladder changes ordering across runtime stacks. The P2 prefill baseline gives 18.7%, 22.2%, and 108.2% test MAPE. GGUF structure helps on all three systems, but fitted efficiencies are not universal.

Smartphone-based Human Activity Recognition (HAR) models often degrade under distribution shifts caused by changes in users, devices, sensor placements, environments, and acquisition protocols. Domain Generalization (DG) addresses this problem by learning from source domains without access to target data. Existing DG methods span training objectives, representation initialization, and architectural modifications, but these components are typically evaluated in isolation despite operating at different stages of the learning pipeline. We present a large-scale controlled benchmark of DG for smartphone-based HAR, comprising more than 410,000 experiments across four model architectures, thirteen training objectives including Empirical Risk Minimization (ERM), five initialization strategies, four architectural configurations, and two shift scenarios: cross-dataset and cross-position. Results show that individual DG components provide limited and highly conditional gains. Alternative objectives rarely outperform ERM consistently, self-supervised initialization helps in specific settings, and architectural modifications, particularly Dynamic Domain Generalization, provide the clearest standalone improvements. Joint configurations, however, frequently outperform their individual components and exhibit complementary and sometimes super-additive interactions, although gains remain model- and shift-dependent. Class-level analysis shows that the strongest configurations mainly improve difficult, shift-sensitive decision boundaries. Finally, oracle checkpoint analysis reveals substantial unrealized performance: source-validation selection recovers only 53% and 26% of the available oracle gain in cross-dataset and cross-position settings, respectively. Overall, effective HAR domain generalization requires jointly designing DG components and robust model-selection strategies.

A deployed language model may refuse a harmful request in English yet comply with its faithful translation, revealing a cross-lingual safety failure that cannot be characterized reliably by output behavior alone. We formalize this phenomenon through an audited equivalence relation and show that, for a declared quotient, representation, metric, feature dictionary, scoring head, threshold, and contrast model, the resulting safety drift admits an exact linear-algebraic characterization. Specifically, the drift is a cross-Gram functional of the within-fiber contrast; its worst admissible value is a support function, while margin invariance is characterized by an annihilator condition. We introduce an intrinsic calibrated exposure measure, governed by the leverage duality $\chi^2=1/\ell-1$, which separates observed drift into three diagnostically distinct regimes: a reader fault removable by recalibration, an exact correction that is too ill-conditioned to be reliable, and a representation-level collision that no readout-only intervention can remove. Thus, identical observed exposure can lead to fundamentally different remediation verdicts. The framework also extends to cone-valued safety heads. An untied order-swap identity provides a diagnostic for the linear control interface; its calibration-state residual predicts a distinct three-control composition error on unseen states and targets, achieving median Spearman correlation $0.964$, compared with $0.269$ for a static cross-Gram baseline. etc.....

Warning: This submission studies stereotypes and biases, and contains toxic and offensive examples, used for illustration purposes only. Fairness benchmarks such as BBQ have become the de facto standard for fairness evaluation across major model families. We argue that these benchmarks are too easy to support their role: training Qwen 2.5 7B Base with Group Relative Policy Optimization (GRPO) on a single BBQ example, or placing that example in context as a one-shot demonstration for in-context learning (ICL), lifts mean BBQ accuracy from 79.9% to 92.9% and 99.0%, respectively, closing 80% of the gap to its large-scale RLHF counterpart (96.1%) with GRPO, and surpassing it with ICL. These effects generalize across model families. A cross-conditioning analysis shows the improvement is carried by the reasoning traces generated by the model, and one example suffices to elicit a category-agnostic “missing evidence” reasoning pattern. We argue that BBQ-style multiple-choice abstention benchmarks measure a single structural cue, and a model that solves them does not thereby become fair. We call for evaluation suites that cover a broader spectrum of fairness alignment.

Recursive self-improvement is becoming essential for autonomous AI agents, whose progress depends on discovering high-value solutions across complex domains. Effective exploration drives this process, yet managing and improving exploration strategies remains a major bottleneck. Current systems face a fundamental dilemma: fixed strategies fail to adapt as search spaces scale, while online policy optimization must navigate vast meta-search spaces under delayed, expensive feedback from long-horizon rollouts. We introduce Dream-RSI, a framework for scalable, recursively self-improving exploration. A lightweight orchestration layer makes exploration explicit and programmable while leaving the underlying base agent unchanged. Our key insight is that accumulated discovery history can act as a replay simulator over the realized search space. By dreaming within this simulator built from historical discovery trees, Dream-RSI obtains immediate, low-cost off-policy feedback to evaluate and refine exploration policies without repeated, expensive online evaluation. The improved policy is then redeployed online to drive further discovery, continuously expanding the simulator pool in a self-improving loop. Across 9 tasks in 4 domains, Dream-RSI achieves competitive quality and improves discovery efficiency in several settings.

Recent work extends recursive self-improvement (RSI) to agent harnesses for long-horizon coding and terminal tasks, enabling agents to improve execution mechanisms from experience. However, generalizable harness RSI remains challenging. First, evolving harnesses on evaluation benchmarks or their subsets makes it difficult to distinguish reusable improvements from benchmark-specific adaptation. Second, single-trajectory updates can conflate systematic harness deficiencies with instance-specific reasoning and solution details, producing modifications that transfer poorly to unseen tasks. Third, localizing recurring behavioral deficiencies within monolithic harnesses is difficult, while whole-harness optimization can entangle unrelated mechanisms and complicate attribution and validation. We propose ModularRSI, a benchmark-disjoint, contrastive, and modular framework for generalizable harness evolution. ModularRSI contrasts successful and failed trajectories for the same task and aggregates evidence across tasks to identify recurring behavioral deficiencies. It decomposes the evolvable harness into five functional modules: Agent Loop, Tool Use, Observation Management, Context Management, and Task Completion Detection. Each module evolves independently within a restricted modification scope, followed by an integration stage that combines the evolved modules into a unified harness and resolves potential conflicts. To support benchmark-disjoint evolution, we curate 2,000 executable evolution tasks from external sources that are disjoint from downstream evaluation benchmarks. Experiments on TB2.0 and SWE-Bench Verified show consistent improvements on unseen in-domain and cross-domain tasks, with the evolved harness also transferring across different foundation models.

Semantic watermarks for diffusion models embed ownership information into the generative process while preserving perceptual quality, but Gaussian-Shading extraction conventionally requires multi-step diffusion inversion to recover the initial noise. Recent one-step methods show that this cost can be reduced substantially. We study this problem through extended flow matching and conditional regression. The key observation is that, near the high-SNR image endpoint, recovering a useful noise statistic given by the first-step output of the extended flow matching in the high-SNR regime is much simpler than reconstructing the full inverse trajectory, and Gaussian Shading only requires the recovered latent to remain in the correct watermark decision region. Based on this observation, we propose a lightweight, prompt-free extractor that decomposes endpoint recovery into an image-like anchor and a noise-oriented residual, which increases the capability of the model to utilize GPU parallel computation. The resulting method avoids iterative inversion and repeated evaluation of a diffusion-scale U-Net, providing an efficient one-step extraction pipeline with a concise theoretical interpretation. The computational cost of extracting noise is lower than that of both OSI and FARI. The github repo is there: https://github.com/TheLovesOfLadyPurple/RAIN-lightweight-NN-for-one-step-semantic-watermark-extraction

Tool descriptions tell an agent what an action can do, but they do not necessarily reveal whether that action is healthy or suitable at the moment of use. Execution harnesses often possess this information through monitors, traces, and recent failures while the model choosing the next action does not. We call this gap operational blindness. We introduce Agent Harness Awareness, a trust-separated design in which the host supplies a small, fresh, verified state snapshot before a consequential action while retaining measurement, enforcement, evaluation, and final dispatch authority. We test its tools-and-network instance with ToolRoute, a controlled isolation benchmark that holds the task and candidate semantics constant and varies the operational information available to the model. In a 216-decision ablation across GPT-5.6 Sol, Claude Sonnet 5, and Gemini 3.7 Flash, verified tool-health state reduced mean observable policy regret by 96.9% against a same-shape neutral control (4,252 ms to 131 ms) and raised single-dispatch completion from 69.4% to 95.8%. In a separate 27-decision end-to-end replication with live local TCP/HTTP execution, OpenTelemetry client spans, and Toxiproxy-injected faults, verified state reduced mean regret from 8,956 ms to 0 ms (zero excess observable operational cost) and completed the live action in 9/9 model-by-fault cases versus 5/9 for the neutral control. These studies establish the tools-and-network instance of Agent Harness Awareness: selected host telemetry materially improves pre-dispatch operational choice. Runtime state, hardware pressure, economics, retry-aware recovery, and heterogeneous selection remain open empirical extensions of the same architecture.

Large language models (LLMs) are increasingly used as simulated respondents in computational social science, yet accurately reproducing population-specific heterogeneous response distributions remains challenging. Existing methods such as demographic prompting are sensitive to prompt design and increase inference overhead, while parameter-based methods struggle to reuse demographic information efficiently, limiting generalization to unseen combinations. In this regard, we propose PARADE, a composable PARAmetric DEmographic injection framework that represents each demographic attribute value with a reusable and composable LoRA module. Specifically, PARADE separates task-common survey capability from demographic-specific value shifts via a shared Task LoRA and attribute-level Demographic LoRAs, with composition calibration to reduce parameter interference. At inference time, target population models can be constructed on demand by recombining the relevant modules, including demographic combinations not observed during training. Extensive experiments on the World Values Survey and SocioBench across three backbone models show that PARADE achieves strong population-level distribution alignment while maintaining competitive predictive performance, reducing KL divergence by up to 52.8% relative to the strongest baseline.

Any objects in space may encounter singularity due to the inverse-square law of the gravitational force, hindering the analysis and propagation of particle motion in space. Regularization in celestial mechanics removes such gravitational singularities in equations of motion by transforming the time and coordinates. This paper addresses a class of regularization methods known as global regularization for restricted three-body problems (R3BPs), which simultaneously removes the two singularities (e.g., Earth and Moon in cislunar space). It enables analysis, propagation, and prediction of particle motion throughout the domain of R3BPs, which is in stark contrast to regularizing each singularity individually. In particular, this paper develops a formulation for global regularization of the spatial elliptic R3BP (ER3BP) under external perturbations. The presented formulation leverages mathematical tools from quaternionic (hyper-complex) algebra and Hamiltonian mechanics. This approach enables us to (i) elegantly handle the four-dimensional coordinates arising from the dimension-lifting operation for spatial regularization, (ii) systematically derive the coordinate-momentum transformation while preserving the canonical structure and incorporating perturbing forces, and (iii) provide geometric insights for analyzing the regularity of the transformation, resulting Hamiltonian, and dynamical flow. Numerical examples demonstrate the validity of the mathematical results as well as significant numerical benefits of the regularization formulation in perturbed Earth-Moon ER3BP.

The generalized source integral equation (GSIE)approach is extended to enhance the compressibility of moment matrices for essentially-convex homogeneous penetrable scatterers. In the proposed indirect surface integral equation (SIE) formulation, the scattered field representations use a single distribution of generalized sources that radiate only weakly into the scatterer. A separate distribution of conventional sources is used for representing the interior fields. The wave interactions between surface subdomains via the generalized kernel are of a reduced effective dimensionality. The corresponding exterior integral operator moment matrix blocks exhibit enhanced rank deficiency. For lossy scatterers, this secures the low-rank (LR) compressibility of hierarchical matrix structures for the full system. For scatterers in a much slower background medium, significant savings with very delayed asymptotic scaling of the costs can be achieved. The GSIE is formulated for the transverse magnetic (TM) scattering problem. It is validated for representative examples and its superior compressibility is demonstrated.

Ensuring the safety of autonomous driving systems (ADS) in a cost-effective and efficient manner remains a critical challenge. Existing law-guided scenario generation approaches are typically limited to a narrow subset of legal rules, resulting in insufficient scenario diversity, and search-based methods often struggle with large and sparse search spaces. To address these limitations, we propose SLaFE (Semantic Law Modeling and Fuzzing based on Cooperative Evolution), a novel traffic violation scenario generation framework designed to systematically evaluate the safety of ADS. SLaFE harnesses the reasoning capabilities of large language models (LLMs) to convert traffic laws into structured scenario constraints. These scenarios are then optimized via a cooperative evolutionary fuzzing algorithm that explores the parameter space to identify boundary cases likely to trigger abnormal ADS behaviors. We evaluate SLaFE on the Apollo platform within the LGSVL simulator using ten real-world traffic regulations. Experimental results show that SLaFE successfully triggered all 10 types of traffic law violations (10/10), outperforming the best existing method, VioHawk (9/10), while others detected no more than 3. Moreover, SLaFE achieved an average triggering time of 5.1 minutes per law type, significantly faster than VioHawk (9.0 minutes) and other baselines. These results highlight SLaFE's effectiveness in discovering diverse and critical law-violating scenarios for ADS testing.

Large language models (LLMs) are increasingly being employed as variation operators in metaheuristics, generating or modifying candidate solutions, heuristics, or programs inside iterative search loops. This shift reframes variation as a model call conditioned on different types of information. We introduce an operator-level framework with two descriptors: (1) the type of prompt-conditioning information at variation time (Numeric, Symbolic, Linguistic), and (2) artifact persistence, identifying what survives the model call (Transient, Amortized, Transfer). The tutorial shows how to classify, build, and select these operators through a worked build template, a method survey, an evidence table, and a cost-aware decision guide.

Vision-Language-Action (VLA) models are studied mainly in robotics, where visual observations and language instructions are mapped to physical actions. This paper introduces Energy Vision-Language-Action (EVLA), a controlled multimodal benchmark for intent-conditioned residential energy management. EVLA frames battery scheduling as a multimodal trajectory-prediction problem in which an RGB energy-field representation, a numerical operating state, and a natural-language objective are mapped to a 16-step battery-action trajectory generated by a finite-horizon sampling-based reference generator. Source windows are derived from public residential electrical-load data, while electricity price, battery state of charge, indoor temperature, and time of day are generated benchmark metadata. A hidden operating regime is encoded only through energy-field texture, enabling paired visual changes while the explicit numerical state is fixed. Crossing 439,203 retained base windows with three hidden regimes and five language objectives yields 6,588,045 multimodal instances. An initial study evaluates 36 configurations over three training seeds using fixed subsets of 5,000 training, 500 validation, and 500 test instances. In the MobileNet-family comparison, removing processed language increases trajectory mean-squared error from 0.3856 +/- 0.0039 to 0.8628 +/- 0.0001, whereas removing vision yields 0.3843 +/- 0.0013, comparable to the full model. The results show strong asymmetry in modality use: the processed-language pathway is strongly associated with prediction quality, while the current RGB pathway provides no aggregate error advantage. These results characterize the fixed pilot subset and executed protocol rather than full-benchmark training. EVLA provides a controlled setting for studying how semantic intent and latent context influence residential energy-action prediction.

Identifying an unfamiliar sign is difficult when a learner remembers its movement but does not know its meaning or formal feature codes. SignTrace addresses this longstanding reverse-lookup problem through natural-language access to a Chinese sign-language dictionary. The system integrates LLM-based dictionary enrichment, action extraction, dictionary-style rewriting, seven-channel retrieval, and candidate reranking over 6,699 entries. It has been deployed for user trials and has received positive informal feedback. Evaluation on a dictionary-derived benchmark of 500 movement-description queries yields 94.0% Hit@1, 97.4% Hit@9, and a mean reciprocal rank of 0.9540. Reranking increases Hit@1 from 71.8% to 94.0%, while component analyses show the contribution of enriched entry descriptions. Median query-processing time is 13.37 seconds with six concurrent queries. By connecting everyday movement descriptions to documented signs and meanings, SignTrace provides a practical tool for identifying unfamiliar signs. Dictionary-derived wording and prior selection within the benchmark limit generalization to descriptions independently produced by users.

Scientific presentations are more than summaries of research papers. They need to present the work in a coherent sequence, explain the main ideas clearly, and help the audience follow the presentation. We present SlideLab, a training-free multi-agent framework for generating scientific presentations from research papers. SlideLab first plans the presentation narrative, then builds and iteratively refines a shared slide deck using agents for content planning, visual generation, layout refinement, and grounding verification. In a blind human preference study, SlideLab was preferred over both open-source and commercial systems on 77% of papers while using roughly 4 times fewer inference tokens than the strongest open-source baseline. We also introduce ConfArena, an audience-oriented evaluation framework that simulates a conference room and assesses presentations slide by slide. ConfArena matches human system rankings and detects injected presentation problems, including falsified numbers, degraded figures, dropped slides, and shuffled slide order.

The Model Context Protocol (MCP) enables AI agents to discover and call tools, but loading every definition becomes expensive as connected catalogs grow. We present Cartograph, a federated MCP proxy that changes agent-visible tool discovery from $O(n)$ catalog traversal to $O(k)$ progressive disclosure. Cartograph combines three mechanisms: (1) operator-attested capability cards, Ed25519-signed descriptions generated under the deploying operator's control rather than ranked publisher copy; (2) Rift, a three-layer confusable-cluster analysis comprising density clustering, query-margin analysis, and token diagnosis; and (3) two-stage retrieval, which ranks servers before tools. On a 22-server, 374-tool deployment, Cartograph exposes three proxy tools instead of 374 definitions. A 49-query author-constructed benchmark yields R@5 of 0.816, compared with 0.592 for a Jaccard keyword baseline, while a measured top-5 discovery exchange uses 475 tokens rather than 42,450 under the stated full-catalog accounting. Rift identifies 49 confusable clusters, including four HIGH-risk clusters in bootstrap-generated cards. An exploratory comparison of 119 LLM-generated descriptions removes the observed zero-distance cluster but shows that mixing card-generation regimes can reduce R@5. Gateway measurements over ten trials add 5ms mean latency (0.8%) relative to direct stdio MCP calls. Cartograph is complementary to code-execution approaches: it controls which tool descriptions are surfaced and records the provenance of the descriptions used for ranking for each query.

We propose a localized core–tail Fourier–Laguerre frame method for approximation and model problems on the real line. The domain is decomposed into a finite core and two semi-infinite tails. Local Fourier extension is used in the core to resolve nonperiodic, oscillatory, and locally nonsmooth structures, while modulated Laguerre frames are used in the tails so that the Laguerre functions approximate only slowly varying decaying envelopes. The local Fourier extension component also provides two data-driven mechanisms: coefficient-energy indicators for internal edge detection and local frequency indicators for selecting tail modulation centers. We derive error estimates that separate the core approximation error, the modulated Laguerre envelope error, the effect of frequency mismatch, and the finite-tail truncation error. The analysis shows that the tail complexity is governed mainly by the residual phase after modulation rather than by the original carrier frequency. Numerical experiments demonstrate high accuracy for oscillatory, multi-frequency, and derivative-discontinuous functions, and a decaying model problem illustrates that the representation can be combined with differential operators through exact interface constraints.

Large-scale egocentric and whole-body human demonstrations are becoming a primary source of data for embodied intelligence. They record what people perceive and do, but rarely the external specification that gave an action its purpose. In construction that omission is consequential: skilled work is directed at project-specific configurations defined in a design model - configurations not yet present in the environment being observed. A mason's transferable competence is not the geometry of one wall but the ability to realise a new geometry from a specification. We introduce the specification-conditioned demonstration: a synchronised record of the physical state a demonstrator perceives, the intended state supplied to them by an external design, and the behaviour connecting the two. We present OJOx, a capture interface that realises this for construction - delivering design geometry to a headset, anchoring it in the physical workspace, rendering it into a demonstrator's stereo passthrough view, and recording that view synchronously with whole-body and hand motion. We report one fully instrumented session - a 33-component wall laid against a specification that changes while the work proceeds - and check the record against the physical scene through an external camera registered independently of the capture. Recorded sessions remain compatible with existing humanoid retargeting infrastructure and replay onto a Unitree G1 in simulation. The result is a data interface for testing whether embodied policies can learn not merely to imitate demonstrated actions, but to act toward specifications absent from their training experience.

Adapting language models to new domains via continual pre-training raises a basic evaluation problem: if the training corpus overlaps with what the model already knows, performance gains cannot be cleanly attributed to new learning rather than pre-existing knowledge. This matters most for knowledge-intensive, task-light (KHTL) robot deployments - pharmaceutical dispensing, hazardous-material handling, facility-specific protocols, where the physical task is simple but the governing rules are proprietary and safety-critical, and where extensive live testing is costly or unsafe. We introduce ORDER (Ontology-driven Decision-making for Embodied Reasoning), a benchmark built on a fictitious world: a 342,069-token synthetic corpus defining a self-consistent physics that cannot appear in any model's pre-training data. ORDER pairs a 500-question knowledge test (ORDER-BENCH) with a harder compositional task, ORDER-SPATIAL: ordering objects for safe manipulation across both familiar and entirely novel scenes. GPT-4.1 without adaptation scores below chance on ORDER-SPATIAL (Kendall's tau = 0.441), showing its priors actively conflict with the invented physics. After continual pre-training, small models improve substantially on both familiar and novel scenes alike evidence of genuine world-model induction rather than memorization. We then carry this through to a robot pipeline: models that answer the knowledge test well often cannot produce valid, executable plans without a further skill-adaptation stage, after which small, fully offline models outperform GPT-4.1 even when GPT-4.1 is given retrieval access to the same rules (Kendall's tau = 0.848 vs. 0.606), on a full perception-to-execution loop demonstrated on a simulated iiwa7 arm with human-in-the-loop correction. Throughout, ORDER-SPATIAL performance, not knowledge-test accuracy is what predicts real plan quality.

Talking-face (TF) deepfakes are detected unevenly by rPPG-based methods across generators. We study two lightweight visual-only cues, rPPG-derived waveforms extracted by RhythmFormer and lip-region discrete cosine transform (DCT) coefficients, on the seven TF methods of Celeb-DF++ under a subject-independent protocol. In-domain, lip-region DCT matches or exceeds the rPPG-derived 1D ResNet on every method except SadTalker, and Concat fusion reaches AUC 0.891 against 0.824 and 0.827 for the unimodal baselines. Under leave-one-generator-out evaluation the cues split: each transfers clearly better to three held-out methods, and IP-LAP is near chance for both. Concat averages 0.798 but falls below rPPG alone where DCT transfers poorly, so static fusion only partly exploits this complementarity. Lip-region DCT outperforms full-face DCT on six of seven methods. We treat the rPPG-derived signal as an empirical cue and do not claim it is cardiac in origin.

Understanding the design space of streaming video diffusion is essential to exploring its potential for generation quality and computational efficiency. We develop a unified analytical framework that relates model and sampler choices, historical conditioning, execution scheduling, and training strategies. The framework accommodates a broad family of causal context-selection policies and makes their computational dependencies and training-inference alignment explicit. Within this design space, we study three representative policies: clean, same-level, and progressive history. On the full VBench prompt set, same-level and progressive history achieve aggregate scores of 85.24 and 85.60, respectively, compared with 84.45 for the clean-history reference. Long-video comparisons further show improved subject consistency and more coherent motion with progressive history. By allowing multiple denoising nodes to be processed together, progressive-history pipelining achieves $1.57$-$2.83\times$ steady-state DiT speedups under our evaluated conditions. We additionally find that LoRA adaptation of the DMD fake-score network improves generation quality using only 2.15% as many trainable fake-score parameters as full-parameter adaptation. Together, these findings show that fully denoised history is not a prerequisite for high-quality streaming generation and motivate the joint design of historical conditioning, execution, and training.