# Research Papers — 2026-10 · bundle page 48 of 156

> Full-text markdown of papers #2351–2400 of 7799 published in 2026-10, newest first. AI-readable mirrors of the full 7799-paper 2026-10 archive (50 papers per file).

Canonical page: https://aipod.dev/papers/archive/2026-10/page/48
[← Previous bundle](https://aipod.dev/papers/archive/2026-10/page/47.md)
[Next bundle →](https://aipod.dev/papers/archive/2026-10/page/49.md)

### PatTransBench: an ipc-stratified, multi-dimensional benchmark for patent machine translation evaluation

Authors: Xiaowei Su, Xusen Cheng
Published: 2026-10-12
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2610.16538v1

We present PatTransBench, a 12,000-item Chinese-English/Japanese/Korean patent machine translation (PatMT) benchmark with a four-dimensional evaluation protocol. The benchmark comprises 6,000 short-text (sentence/claim) and 6,000 long-text (paragraph) pairs, evenly split across three language pairs and both translation directions, built via IPC-section-stratified sampling and dualthreshold quality gating from upstream patent-family parallel corpora. The evaluation protocol spans literal n-gram overlap (BLEU, ROUGE-L, chrF), cross-lingual semantic textual similarity (CLSTS), 38-probe-language consistency testing, and serving performance under concurrency. A case study on PatMT2-7B and its fp8/AWQ4/AutoRound-4bit variants reveals direction-dependent quantization effects masked by aggregate scores and non-monotonic language-failure patterns across quantization schemes that are invisible to single-metric evaluation. We plan to release the benchmark, protocol, and case-study results upon publication to support reproducible PatMT evaluation.

---

### Learning Graded Sparse Mechanisms: A Representation Theorem and Certified Generalization

Authors: Mark Kozdoba, Shie Mannor
Published: 2026-10-12
Categories: cs.LG, cs.MA, stat.ML
arXiv: https://arxiv.org/abs/2610.16537v1

In a variety of multi-agent and physical environments, an element's update depends strongly on a few other elements and weakly on many, and both the identity and the number of the elements that matter change with the state of the system. Such graded sparse mechanisms are common, for instance, in flocks, crowds, learned particle simulators and the neighborhood aggregation of graph networks, yet they have received little systematic study as a class of functions with the grading given. In this paper we study the supervised learning of the mechanism when the grading is provided by the environment, as a weight on every element. First, we define the class by axioms, where each element's influence is capped by its weight, the dependence on the weights is regular, and no element is distinguished, and prove a representation theorem: the class is exactly the Lipschitz ball of a transport metric on weighted multisets, and the metric is forced by the axioms. Next, the geometry of the representation gives generalization bounds that depend on the data in two ways. One is via the sparsity profile of the data -- the law of how many elements matter and by how much -- which bounds the rate over all designs with that profile, with an exponent from the profile and no dependence on the number of elements. In a number of natural cases we also show matching lower bounds, up to constants in the exponent. The other is empirical: due to the representation result, classical empirical discrepancy bounds apply to the class as they stand, and the resulting bound is the transport distance between the two halves of the sample in the class's own metric, computed exactly by a matching, with an additional concentration term. Finally, we study a simulated Vicsek flock under four interaction rules at two noise levels, and show that the certificate is non-vacuous, a small fraction of the trivial value on every rule.

---

### Learning, Compression, and Meta-Complexity

Authors: Halley Goldberg, Mikito Nanashima, Igor C. Oliveira
Published: 2026-10-12
Categories: cs.CC, cs.LG
arXiv: https://arxiv.org/abs/2610.16535v1

The idea that learning is closely tied to compression has long served as both a powerful principle and, at times, an informal analogy across machine learning and learning theory. Here we focus on settings in which that intuition can be made mathematically precise, surveying formal connections among learning algorithms, data compression, and computational complexity theory, with an emphasis on recent progress and current frontiers. We describe how techniques from complexity theory can guide the design of both supervised and unsupervised learning algorithms with provable guarantees, and conversely, how learning algorithms can yield new insights and results in complexity theory. A central theme is the role of data compression as a bridge between learning and complexity, especially through an area called meta-complexity, which studies the computational difficulty of compression problems and time-bounded variants of Kolmogorov complexity.

---

### Does the Explanation Drift Before the Guarantee? Label-Free Early Warning on Evolving Graphs

Authors: Tomasz Szandała, Damian Raczkowski
Published: 2026-10-12
Categories: cs.AI
arXiv: https://arxiv.org/abs/2610.16534v1

Deployed models can fail silently as the world they describe evolves. Attribution drift has served as an early warning in tabular data; we ask whether, in graph streams, a model's own explanations can forewarn the loss of nominal conformal coverage. Doing so requires repairing the naive drift statistic twice. A vanishing edge loses mask mass regardless of whether the model relied on it, so we condition the importance on edge presence. Evolving graphs also break edge identity: a relabelling that changes nothing shifts the identity-based statistic by 0.69 on synthetic graphs and 0.83 on real ego-networks, whereas aggregating over structural roles brings both to 0.0 and stays well-defined when the support changes. We test six pre-registered regimes on thirty seeds each. A two-channel monitor, tracking conditional mask mass and its distribution over roles, precedes coverage collapse by 8 to 14.5 steps across four gradual regimes (node addition, removal, replacement, and edge thinning), all surviving Holm correction. A sudden knockout yields the expected zero lead, and a harmless change triggers no alarm. Leads track the length of the transitional phase (Spearman 0.79 over 150 runs). On real transaction ego-networks, the untuned mass channel forewarns staged drift by 27 steps and is the only channel, attributional or structural, that never alarms on a stationary stream. The main limitation is that the mass channel is blind to covariate drift, which breaks the guarantee in every run; the distributional channel detects it and also rescues the rewiring regime, so we report the two channels separately rather than as a union.

---

### Spectral-Driven Adaptive-Rank Optimizer for Large Language Model Reinforcement Learning with Verifiable Rewards

Authors: Zhipeng Chen, Wayne Xin Zhao, Ji-Rong Wen
Published: 2026-10-12
Categories: cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2610.16533v1

Reinforcement learning with verifiable rewards (RLVR) is reshaping how large language models (LLMs) acquire reasoning capabilities, yet the optimizers used remain largely inherited from teacher-forcing training. This raises a fundamental question: does RLVR induce a distinct optimization geometry that calls for a different optimizer? We investigate this question through the spectral structure of LLM gradients and find that gradient energy progressively concentrates into a shrinking effective subspace, which is characteristic of RLVR, as it is absent under supervised fine-tuning (SFT) and is largely removed when the verifiable reward is replaced by a random one. Moreover, the rank required to capture a fixed fraction of gradient energy steadily decreases over training, revealing that the effective dimensionality of useful updates itself evolves with training. These findings motivate a new optimization principle: rather than prescribing the update rank externally, an optimizer should adapt its rank to the evolving spectral structure of the gradient. We introduce Ares (Adaptive Rank-Effective Spectral Optimization), which estimates the effective update rank from the gradient's spectral energy distribution, tracks its evolution, and confines the gradient signal driving each update to the resulting low-rank subspace. Across LLMs of different scales, Ares improves accuracy over widely used optimizers and generalizes across RLVR algorithms and tasks. Results suggest that effective optimization may require optimizers to be designed around the geometry and dynamics of the training process. We release the code and data at https://github.com/RUCAIBox/Ares.

---

### Support or Magnitude? Characterizing Neuron Relevance for Instruction Data Selection

Authors: Yifan Le
Published: 2026-10-12
Categories: cs.CL
arXiv: https://arxiv.org/abs/2610.16531v1

Model-internal signals, such as neuron activations and gradients, have been used for instruction data selection, yet how neuron relevance should be characterized remains underexplored. In particular, should selection prioritize the magnitude of relevance or how broadly it is distributed across neurons? To investigate this question, we construct sample-level relevance profiles using AttnLRP-guided attribution and introduce two complementary criteria: Effective Relevance Support (ERS), which measures the effective support of positive relevance, and Positive Relevance Magnitude (PRM), which captures its average strength. Our analyses reveal that these characterizations produce distinct sample rankings and selection preferences despite sharing the same underlying relevance profiles. On Qwen2 models, PRM exhibits stronger associations with activation-based selection, whereas ERS produces substantially different rankings. These differences extend to task composition, within-task sample preferences, and semantic selection patterns, with varying degrees of agreement across backbones. Experiments across three language models and five downstream benchmarks demonstrate the practical value of relevance-guided selection, with ERS achieving the best overall average performance rank among the evaluated methods. Our findings highlight that how neuron relevance is distributed and characterized, rather than its magnitude alone, matters for instruction data selection.

---

### Deep Learning-based Mechanical Characterization of Synthetically Generated Plant Fibers under Transverse Compression

Authors: Nikolaos Kegkeroglou, Bilal Sabillah, Myrto Inglezou, Vincent Placet, Céndric Clevy, Antonios Porichis
Published: 2026-10-12
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.16529v1

Plant fibers are increasingly used as sustainable reinforcements in composites, but the intrinsic variability of their microstructure causes substantial scatter in transverse mechanical behavior, complicating composite design. Characterizing this behavior relies on finite element (FE) simulations requiring several minutes of high-performance computing per fiber, making high-throughput characterization across realistic fiber populations impractical. In this study, we propose a deep learning framework replacing this FE bottleneck with a fast pipeline. It includes two coupled recursive U-Nets predicting the step-by-step evolution of fiber deformation and internal stress fields under transverse compression, followed by a regression CNN estimating the transverse Young's modulus from the final stress field. Unlike prior approaches predicting stress fields for fixed configurations or idealized geometries, our framework recursively resolves the full temporal stress evolution for synthetic fibers derived from real hemp cross-sections, with or without a lumen. Trained on 1,000 synthetic fiber geometries with FE ground truth, it reproduces the stress field sequence with a Structural Similarity Index Measure (SSIM) of $0.9574 \pm 0.0074$ and a Root Mean Squared Error (RMSE) of $0.0345 \pm 0.0079$. It estimates transverse Young's modulus with an $R^2$ of $0.938 \pm 0.024$, a Mean Absolute Error (MAE) of $24.31 \pm 1.23$ MPa, and a Mean Absolute Percentage Error (MAPE) of $5.25 \pm 0.91\%$. Inference requires under one second per fiber versus five minutes per FE simulation, drastically reducing computational cost. This offers a fast, reliable tool for mechanical characterization of synthetic plant fibers, with promising applications in designing natural fiber-reinforced composites.

---

### Reinforcement Learning with Hawkes-Structured Networks for Market Making

Authors: Fernando Acero, Philipp Plank, Yufei Zhang
Published: 2026-10-12
Categories: cs.LG, q-fin.CP
arXiv: https://arxiv.org/abs/2610.16524v1

Market making under realistic order-flow dynamics is challenging due to the clustered, self-exciting nature of order arrivals, which are naturally modeled by non-Markovian Hawkes processes. We propose a model-free reinforcement learning (RL) framework for market making with general Hawkes order-flow dynamics. We first establish non-asymptotic performance bounds that quantify the loss incurred by approximating the infinite-dimensional Hawkes state using a finite collection of exponentially weighted features. Guided by this analysis, we introduce the Hawkes-Structured Network (HSN), a novel neural architecture that encodes historical order flow through a learnable mixture-of-exponentials filter. HSN is trained end-to-end using model-free RL, jointly optimizing the feature representation and quoting policy directly from reward feedback. Experiments on single- and multi-asset market-making problems with mixture-of-exponentials, power-law, and regime-switching kernels demonstrate that HSN outperforms general-purpose neural architectures, achieving higher trading rewards while implicitly recovering the underlying order-flow intensities.

---

### EAU: Environment-Agent-User Co-evolution for Training User-Centric Agents

Authors: Min Cai, Dan Zhang, Gao Xing, Libin Ma, Zhaohui Xu, Xi Ye
Published: 2026-10-12
Categories: cs.AI, cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2610.16523v1

Training LLM agents to interact with users over multiple turns is challenging because the learning signal depends not only on the agent, but also on the user simulator and environment, which jointly shape the agent's experience. Simulated users can be stochastic or inconsistent, while environments may contain unreliable tools, incomplete state coverage, and incorrect verifiers. Prior work mainly addresses the resulting instability through reward shaping or denser supervision, while treating these components as fixed. We instead introduce EAU, a meta-agent framework for environment--agent--user co-evolution. During training, EAU monitors trajectories, tool traces, rewards, and training statistics; diagnoses failures and uninformative interactions; and proposes targeted modifications to the user simulator, environment, reward design, and RL configuration. Validated changes are hot-swapped into the ongoing run, allowing the training setup to evolve together with the policy. These adaptations also persist across runs, so later training can build on previously discovered improvements rather than restarting from a fixed setup. We enable this process through an editable symbolic representation of user and environment dynamics, while the agent remains a neural LLM optimized with RL. Across three interactive-agent benchmarks and two backbone models, EAU consistently outperforms prior RL methods, improving the overall score over the strongest baseline by 14.7--15.0% and increasing successful task completion on ColBench by up to 36.7%, showing that co-evolving the training system with the policy can substantially improve multi-turn RL.

---

### Learning Reachability-Based Safety Representations for Real-Time Robot Navigation in Unknown and Dynamic Environments

Authors: Bojan Derajić
Published: 2026-10-12
Categories: cs.RO, cs.LG, eess.SY
arXiv: https://arxiv.org/abs/2610.16522v1

Safe robot navigation in unknown, dynamic environments requires more than geometric collision checking: a robot still far from an obstacle may be unable to avoid it because of its speed, orientation, or the obstacle's motion. Hamilton--Jacobi (HJ) reachability analysis captures this in a value function whose sign separates states that admit a safe continuation from those where collision is unavoidable, but computing it online for every newly observed environment is too expensive. This dissertation trains neural implicit representations offline on reachability solutions to give optimization-based controllers this information in real time. Since the value function never exceeds the constraint, it equals a directly computed constraint, such as a signed distance, minus a nonnegative dynamics correction. Learning only this correction keeps represented obstacles outside the predicted safe set despite approximation errors. The first two of five methods use the learned function as a terminal constraint in model predictive control, for static local environments and predicted obstacle motion. The next two use it as a barrier candidate in control barrier function safety filters, for static environments and, through pairwise barriers in relative coordinates, for a varying number of moving obstacles. Simulations and hardware experiments on ground and aerial robots show better collision avoidance than geometric and hand-designed baselines, especially for short horizons and moving obstacles. The fifth method replaces the grid-based offline solver, which limits the first four to low-dimensional models. For control-affine dynamics with polytopic inputs, the HJ Hamiltonian is maximized at a vertex, so a parallelizable tree search over vertex inputs can generate training labels. With a loss from the HJ variational inequality, these labels yield barrier candidates for systems with up to 13 states.

---

### Asymptotic Risk of Learning-Rate Schedules for Least-Squares Regression with Discretely Distributed Features

Authors: Noé Colomban, Francis Bach, Fabian Schaipp
Published: 2026-10-12
Categories: cs.LG, math.OC, stat.ML
arXiv: https://arxiv.org/abs/2610.16518v1

We study the last iterate of stochastic gradient descent (SGD) on streaming least squares with discretely distributed features (Slock's model), which decouples the dynamics across eigendirections. Under power-law source and capacity conditions, we derive the exact finite-time risk and asymptotic equivalents of constant and Warmup-Stable-Decay (WSD) schedules. With a tuned base learning-rate, the resulting scaling law decays as $Θ(T^{-q})$, where $T$ is the horizon, and $q$ is independent of the cooldown length $c \in (0,1)$ but is strictly smaller for the constant schedule: any cooldown of fixed relative length $c > 0$, however short, improves the risk by a polynomial factor, while its length only affects the constant, which is minimized by linear decay ($c = 1$). Finally, we show that the optimal risk depends on the batch size $b$ and horizon $T$ only through $bT$, up to a critical batch size.

---

### Trust-Calibrated VLA Policy Post-Training with Evolving Imagination

Authors: Wenzhan Li, Yiran Qin, Heng Zhou, Huirui Wang, Yulan Guo, Ruimao Zhang
Published: 2026-10-12
Categories: cs.RO, cs.LG
arXiv: https://arxiv.org/abs/2610.16517v1

Vision-Language-Action (VLA) models are often pretrained via imitation learning, making reinforcement learning (RL) post-training important when deployment moves beyond demonstration coverage. Existing refinement paradigms expose a reliability--efficiency tension: model-free refinement avoids simulator bias but is interaction-intensive, whereas model-based refinement can exploit unreliable imagination even as its simulator adapts. We propose Model-Evolving Reinforcement Learning (MERL), a trust-calibrated framework that couples policy optimization with an online-updated simulator comprising a world model and a reward proxy head. Across stages, simulator reliability controls the imagined-data ratio and rollout horizon; within rollouts, chunk-level trust controls replay selection and optimization weight. For recursive imagination without exact grounded continuations, a residual predictor estimates trust without querying future ground truth. On perturbed LIBERO-PRO tasks, MERL reaches 79.7% average success, improves high-threshold learning speed by 10.9% over the matched model-free control, and outperforms same-protocol external baselines across three training seeds. Fixed-data adaptation from 50 demonstrations per task achieves 80.0% average success on two physical-robot tasks.

---

### Intent-Triggered Anchoring for Multi-Token Format Units in Parallel Decoding of Diffusion Language Models

Authors: Junhyeok Lee, Kyu Sung Choi
Published: 2026-10-12
Categories: cs.CL
arXiv: https://arxiv.org/abs/2610.16516v1

In parallel decoding, diffusion language models commit several tokens per forward pass, so co-committed tokens cannot see each other, breaking multi-token format units that parsers need. On ECHO, an open-weight one-step chest X-ray (CXR) report generator, 32.7-36.7% of evaluation reports lack a parseable Impression heading and score as empty. We propose intent-triggered anchoring, a training-free decoding rule that completes a format unit once the model commits its first token, inserting only units with a unique, content-free completion; fallbacks write the unit when the model stops or loops without starting it. On three CXR evaluation sets, the rule raises readable rates from 63.3-67.3% to 100.0% with no regressions and a 95th-percentile latency ratio of 0.68-1.28 against official decoding. Broken headings state findings absent from the reference; adding the heading at the end leaves them in 15.5-31.0% of reports, our rule in at most 0.6%. For math answer markers, the rule raises accuracy in the most parallel setting of five diffusion language models by 5.8-20.8 points on GSM8K and 5.1-17.1 on GSM-Symbolic, held out from tuning. With lenient answer extraction, gains are 3.3-12.5 and 2.7-6.3 points; the rule adds 1.4-8.6 points over a stop-only fallback for four of the five models.

---

### Boosting the Learning for Ranking Patterns

Authors: Nassim Belmecheri, Noureddine Aribi, Nadjib Lazaar, Yahia Lebbah, Samir Loudni
Published: 2026-10-12
Categories: cs.AI
arXiv: https://arxiv.org/abs/2610.16511v1

Pattern mining is a valuable tool for exploratory data analysis, but identifying relevant patterns for a specific user is challenging. Various interestingness mea- sures have been developed to evaluate patterns, but they may not efficiently estimate user-specific functions. Learning user-specific functions by ranking patterns has been proposed, but this requires significant time and training samples. In this paper, we present a solution that formulates the problem of learning pattern ranking functions as a multicriteria decision making problem. Our approach uses Analytic Hierarchy Process (AHP) to elicit weights for different interestingness measures based on user preference. We also propose an active learning mode with a sensitivity-based heuristic to minimize user ranking queries while still providing high-quality results. Experiments show that our approach significantly reduces running time and returns precise pattern ranking while being robust to user error, compared to state-of-the-art approaches.

---

### Locally adaptive step sizes for Hamiltonian Monte Carlo methods

Authors: Jimmy Huy Tran, Tore Selland Kleppe, Nawaf Bou-Rabee
Published: 2026-10-12
Categories: stat.CO, stat.ME, stat.ML
arXiv: https://arxiv.org/abs/2610.16506v1

Exact reversible locally adaptive step sizes are introduced by Bou-Rabee et al. (2026) in the context of the No-U-Turn Hamiltonian Monte Carlo (HMC) sampler for the purpose of high performance HMC sampling in challenging target distributions. This paper carefully explains and illustrates the locally adaptive step size procedure under more general assumptions. Alternative error measures are introduced with the aim of further enhancing MCMC performance by reducing the prevalence of reversibility failure rejections. In addition, the paper shows how to introduce the locally adaptive step sizes also within standard HMC and Metropolis-corrected kinetic Langevin/generalized HMC (GHMC). Finally, a numerical study of the relative merits of the proposed error measures in the context of different Hamiltonian flow-based MCMC methods is provided.

---

### pi-Net: Empirical Latent Representations for Differentiable Physics Heads in Data-Scarce Systems

Authors: Akshay Dhalpe, Arnab Chattopadhyay, Shahriar Bakrani Balani, Eric Cortanéa, Sasu Tarkoma
Published: 2026-10-12
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2610.16504v1

Developing reliable predictive models for multi-physics systems often forces a compromise between computationally demanding physical simulators and data-hungry, opaque machine learning (ML) algorithms. While incorporating dimensional analysis into directed computation graphs offers a structurally sound alternative to explicitly solving partial differential equations (PDEs), such formulations typically yield underdetermined systems that act as imperfect simulators. We introduce pi-Net, a physics-informed latent variable model that resolves these imperfect simulators. The hybrid pi-Net architecture consists of an encoder that maps inputs into physically normalized latent space, which are then decoded into final predictions by a differentiable physics head. Evaluated on a real-world, data-scarce Gas Metal Arc Welding (GMAW) case study, the framework achieves an R2 of 0.864 with only 60 experimental samples. The results demonstrate that pi-Net significantly outperforms purely data-driven baselines in data-scarce regimes while providing causal, process-level explainability.

---

### KARAMEL: Knowledge-Aware Ranking for Multimodal Entity Linking in the Arts via Heterogeneous Graphs

Authors: Raffaele Scaringi, Alejandro Sierra-Múnera, Gennaro Vessio, Giovanna Castellano, Ralf Krestel
Published: 2026-10-12
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2610.16502v1

Discovering and disambiguating entities from multimodal and heterogeneous data is a key challenge in discovery science, where relevant knowledge is often implicit and distributed across modalities. We present KARAMEL, a framework for multimodal entity linking (MEL) - the task of aligning visual and textual mentions to entities in a knowledge base - in the cultural heritage domain. Unlike prior vision-language models that primarily rely on surface-level multimodal alignment, our approach incorporates sociohistorical and contextual knowledge through a knowledge-aware architecture that fuses a vision-language encoder with a heterogeneous graph neural network. This design enables contextual reasoning over complex relational structures and supports robust entity disambiguation under ambiguous or underspecified evidence. To benchmark multimodal discovery in this setting, we introduce WikiMuSA (Wikidata-based Multimodal Semantic data for Art), a new dataset that links artworks to Wikidata entities via images, textual descriptions, and structured contextual knowledge. Experimental results show that KARAMEL surpasses strong baselines, particularly on implicit or abstract mentions, while ablation studies highlight the critical role of domain-specific, structured context in learning expressive representations. While instantiated in the cultural heritage domain, KARAMEL addresses a general knowledge discovery problem and applies to other scientific settings involving multimodal observations and heterogeneous background knowledge. Both the dataset and source code are made publicly available.

---

### Self-Verifier Capture in Unified Visual Generators: Diagnosing and Repairing Closed-Loop Self-Reward

Authors: Binyan Chen, Tong Pan
Published: 2026-10-12
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.16500v1

Image-space reward requires decoding and a separate judge; unified visual generators can instead reuse their understanding pathway for decode-free, self-supplied reward. Optimizing against this verifier closes the self-reward loop; shared trainable weights let one update move generation and scoring even with the reward gradient stopped. We call the resulting failure self-verifier capture: self-reward rises while independently measured quality falls. Matched semantic/denoising by shared/frozen cells on Scale-RAE isolate two routes: the policy exploits on-manifold blind spots of a frozen semantic judge, while shared updates additionally rewrite the judge. A frozen judge also becomes stale, and repeated self-refresh yields progressively weaker recoveries. The diagnosis yields a two-part repair: locking the warmed reward read blocks rewriting, while sparse, parameter-disjoint external anchors counter residual exploitation. Unanchored rollouts retain semantic reward across multiple latent noise levels, providing dense latent reward that terminal image-space judging cannot supply, while a drift-triggered controller allocates anchors to high-drift rollouts and propagates their calibration within each group. At matched budgets, the controller improves on fixed-dose anchoring, increasingly so as coverage tightens. Both the unanchored failure and the repair transfer to the higher-performing, architecturally distinct Show-o2. Across both bases, the complete method significantly outperforms a construct-matched every-rollout judge on GenEval and DPG-Bench while externally scoring only 1/16 of rollouts. At 1/32 coverage, it scores higher in all four comparisons and remains jointly non-inferior.

---

### From Caricature to Collective Artifact: How Persona Construction Drives Polarization in Generative Social Simulations

Authors: Dimitris Michailidis, Petter Törnberg, Fernando P. Santos
Published: 2026-10-12
Categories: cs.SI, cs.AI
arXiv: https://arxiv.org/abs/2610.16498v1

Generative Social Simulation (GSS) leverages Large Language Models (LLMs) to create sophisticated personas that define agents with demographics, emotions, and political opinions. While prior work shows that personas compress outputs into stereotypical caricatures, there is no comprehensive study of how these caricatures evolve into collective artifacts when they interact. Using a Twitter-like GSS platform with personas derived from the American National Elections Study (ANES), we ablate three layers of partisan identity: affective polarization markers, party identification, and vote choice, while retaining demographics, ideological positioning and important issues. We find that collective outcomes depend heavily on how survey responses are translated into personas. Removing a few explicitly partisan statements moves the network from near-complete segregation to cross-party ties approaching those of real Twitter networks. We argue that matching one outcome is not evidence for behavioral validity and point to salience as a possible explanation. Surveys ask for a position on every question regardless of its importance to the respondent, and persona construction turns each answer into a direct behavioral instruction. We conclude with a checklist for more comprehensive evaluation of personas in future works.

---

### Drafting Agents, Not Just Tokens: Speculative Recursive Self-Improvement for Agentic Systems

Authors: Xin Xu, Ke Cheng, Hao Wang, Yangru Huang, Yanghao Xiao, Zhichao Chen, Zhisheng Zhong, Xun Liu, Yi Li, Qian Li, Haibin Lv, Jun Zhang et al.
Published: 2026-10-12
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2610.16491v1

Improving an agent through successive revisions requires balancing exploration and evaluation under a limited budget. A narrow search can miss useful sequences of edits, while fully evaluating every intermediate version can exhaust the budget before those sequences yield a verified improvement. We introduce Speculative Recursive Self-Improvement (SpecRSI), which drafts executable successor agents before their predecessors are fully evaluated, while requiring complete verification before adoption. Verification-Aware Tree Search (VATS) allocates the budget between further edits and outstanding checks using the predicted verified return of a continuation. Group Verification Optimization (GVO) learns the improver from completed searches while keeping the real evaluator fixed. In the main scientific comparison, terminal gains exceed Janus by 33.3%--40.0%. Additional evaluations show that SpecRSI outperforms Janus in all 50 scientific backbone--task pairs and exceeds MetaRSI-v1 by 1.85%--2.38% in macro-average system score across twenty backbones. Coding-validation success reaches 69.4% versus HGM's 64.3% at a 20.8% lower target-cost index.

---

### Communication, Vocalisation, and Language

Authors: Peter Wallis
Published: 2026-10-12
Categories: cs.CL
arXiv: https://arxiv.org/abs/2610.16490v1

The contention is that the current batch of Large Language Models don't *understand* language, but they *do* it. Rather than LLMs forming some kind of ``distributed representation'' of symbols, the claim is that LLMs are capturing ``Cultural Intelligence'' and grounding symbols in what the sociologists call practices.

---

### Growing and Transferring Pathways for LLM Reinforcement Learning

Authors: Haolin Tian, Yikang Li, Tonghan Wang
Published: 2026-10-12
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2610.16489v1

Reinforcement learning (RL) can substantially improve the reasoning capabilities of large language models, but an effective training run requires considerable resources. Reusing the computation acquired through RL across models would extend the value of that training. We propose GrowRL, an RL method that organizes learned updates into compact computational pathways through structural growth. GrowRL freezes the pretrained backbone and adds an independent causal attention branch and a feed-forward network branch at each Transformer layer. Zero-initialized output projections preserve the pretrained policy at initialization, and RL updates only the added pathways. The learned pathways remain explicit after training, allowing transfer to structurally compatible target models and subsequent adaptation with the target backbone frozen. The architecture also supports sequential transfer of additional pathways. By keeping the learned attention and feed-forward computation separate from the pretrained weights, GrowRL provides an explicit structure for preserving and reusing the outcome of an RL run. Experiments across five backbones show competitive mathematical reasoning and out-of-distribution performance with a small number of trainable parameters. Learned pathways provide immediate gains across all three transfer directions, with performance after only 10 target-side updates already close to that after 100 updates in every case. These results support computational pathways as parameter-efficient and reusable units of reinforcement learning adaptation.

---

### SLEM: Spectrum-Aware Latent Expert Merging

Authors: Dohyeon Kim, Bedionita Soro, Bruno Andreis, Sung Ju Hwang
Published: 2026-10-12
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2610.16484v1

Mixture-of-Experts (MoE) models scale model capacity efficiently by activating only a small subset of experts for each input, while still requiring all parameters to be stored in memory, resulting in substantial memory overhead. Expert merging techniques have recently emerged as a promising compression paradigm that reduces this overhead by consolidating multiple experts into a smaller set. However, existing approaches have two key limitations: (i) direct convex interpolation in parameter space is not generally guaranteed to yield meaningful interpolation of expert functions; and (ii) estimating expert similarity or merging coefficients often relies on calibration data, introducing dependence on the calibration distribution. To address these limitations, we propose Spectrum-Aware Latent Expert Merging (SLEM), a calibration-free expert merging framework that occurs in a structured latent representation of expert weights using a VAE. SLEM groups experts within and across layers based on latent similarities and derives merging coefficients from expert weight spectra. It then performs geometry-aware latent interpolation and decodes the resulting representations into merged expert weights. Compared with existing expert merging methods, SLEM achieves comparable or better performance on general knowledge benchmarks and higher average performance on generation benchmarks involving mathematical reasoning and code generation.

---

### MiMaAlign: Hierarchical Micro-Macro Alignment for Physically Plausible Video Generation

Authors: Yansong Liu, Litong Gong, Weijie Li, Tiezheng Ge, Ming Zeng
Published: 2026-10-12
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.16482v1

Text-to-video (T2V) diffusion models have made remarkable progress in generating high-fidelity visuals, yet their outputs still often exhibit physically implausible phenomena, including discontinuous dynamics and implausible event interactions. Existing representation alignment methods mainly transfer generic video features or short-range motion cues, treating physical plausibility as a single-level representation problem. To address this issue, we propose MiMaAlign, a hierarchical micro-macro alignment framework that transfers complementary dynamics-oriented and event-oriented knowledge into different abstraction depths of a pretrained video diffusion model. Specifically, micro-level continuous dynamics alignment imposes constraints by aligning diffusion features with a continuous-time motion representation learned by a Neural Ordinary Differential Equation (Neural ODE), while macro-level event-commonsense alignment provides representations of object interactions and event semantics through an instruction-prompted vision-language model (VLM). With training-only auxiliary objectives, we align the complementary supervision from the two teachers with shallower and deeper generator features, respectively, enabling the generator to internalize the corresponding physical priors. Empirical evaluations show that, on CogVideoX-2B, MiMaAlign achieves a relative improvement of up to 21.0% in the Physical Commonsense (PC) score on the VideoPhy benchmark over the baseline, and also delivers substantial gains on the highly challenging VideoPhy2 benchmark, while maintaining general video quality on VBench and being supported by blind user comparisons.

---

### PXRD-Bench: How Close Are We to Determining Crystal Structures from Powder Diffraction?

Authors: Dongyeop Woo, Nofit Segal, Nayoung Kim, Minkyu Kim, Rafael Gomez-Bombarelli, Sungsoo Ahn
Published: 2026-10-12
Categories: cs.AI, cond-mat.mtrl-sci
arXiv: https://arxiv.org/abs/2610.16480v1

Determining crystal structures from powder X-ray diffraction (PXRD) would enable routine characterization of newly synthesized materials, including those absent from reference databases. Generative models offer a promising approach, but existing evaluations remain fragmented and largely assume single-phase samples with known composition. To fill this gap, we introduce PXRD-Bench, a unified benchmark that extends evaluation to unknown per-element atom counts and recovery of both structures in two-phase mixtures. It provides approximately 4.1 million training structures with simulated PXRD, tests generalization to unseen compositions and structural prototypes, and evaluates recovery from experimental PXRD after excluding structurally matching training entries. To provide a strong reference baseline, we also introduce Bragglite, a PXRD-to-structure model developed through a systematic study of PXRD modeling choices. It achieves a 44.2\% Top-1 match rate on experimental RRUFF patterns with known composition, 24.6\% above the strongest prior baseline. Further analysis yields several key findings: some models fail to benefit from experimental PXRD, while generalization to unseen structural prototypes and recovery from two-phase mixtures remain challenging. Together, our benchmark, systematic studies, and Bragglite clarify the capabilities and limitations of current PXRD-to-structure models.

---

### XAI-Guided Fourier Domain Generalisation for Pure-Simulation Sim-to-Real Deep Reinforcement Learning

Authors: Wei Yao, Chuyao Wang, Ziwei Wang, Nabil Aouf
Published: 2026-10-12
Categories: cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2610.16478v1

Deep reinforcement learning has shown strong potential for vision-based autonomous driving, but policies trained in simulation often suffer severe performance degradation when deployed in the real world. Existing sim-to-real methods commonly address this gap through image-transfer-based domain adaptation or policy-agnostic domain randomisation. However, the former requires laborious acquisition of target-domain data and is often limited to the deployment domain, whereas the latter introduces unguided perturbations that may not reflect the underlying sim-to-real discrepancy, leading to suboptimal transfer performance. To overcome these limitations, we propose XFD, an explainable artificial intelligence (XAI)-guided Fourier domain generalisation framework for pure-simulation sim-to-real reinforcement learning. The proposed framework extracts explainability attribution maps from simulated observations, and converts these attributions into policy-conditioned Fourier-domain interventions. By perturbing appearance-sensitive spectral components while preserving task-defining spatial structures, XFD constructs robustness-enhancing visual domains entirely within simulation. Subsequently, the proposed robust policy optimisation guides the policy towards attending to task-critical features while suppressing sensitivity to irrelevant visual information. Extensive experiments in simulation and zero-shot real-world deployment demonstrate that XFD consistently improves traveled distance and reward over standard policy and randomly domain-randomised baseline, especially on unseen maps with cluttered backgrounds. Attribution-based analysis further proves that the proposed policy preserves more consistent task-relevant visual evidence across simulation and real-world observations.

---

### Ontology-Based Post-Coordinated Expression Assembly: from Clinical Narrative to Queryable Knowledge Graphs

Authors: Johann Pignat, Christophe Gaudet-Blavignac, Julien Ehrsam, Jean-Philippe Goldman, Cyrille Duret, Phil Cheng, Mina Bjelogrlic, Michel Cuendet, Christian Lovis, Olivier Michielin
Published: 2026-10-12
Categories: cs.CL
arXiv: https://arxiv.org/abs/2610.16474v1

Clinical oncology documentation records decisive information in free text, where it remains readable but not retrievable at scale, leaving cohort identification dependent on incomplete diagnostic codes or manual review. The current standard, coding individual terms, captures what a document mentions, not what it asserts: meaning lies in how those terms relate to one another. We present a method for deriving queryable knowledge graphs from clinical narratives, conforming to the Extended Grammar of SNOMED CT across multiple ontologies. The method specifies how entities should be annotated and related. It defines five relation classes, each bound to a construction rule that assembles the entities and relations into post-coordinated expressions. A deterministic implementation, UNIGRAPHE, applies those rules to construct knowledge graphs from annotated text. Applied to expert-annotated real-world oncology documentation, UNIGRAPHE reconstructed 5,299 annotations into 2,196 post-coordinated expressions and shows that relation structure is the missing piece to make clinical narrative formally queryable.

---

### Auditing Observation-Quality Shortcuts in the Evaluation of Person-Centered Motion for Video Anomaly Detection

Authors: Chaehyun Kim, Changwon Lim
Published: 2026-10-12
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.16471v1

Video anomaly detection (VAD) identifies deviations from normal patterns, but such deviations do not necessarily correspond to safety-critical events. We focus on human-centered emergencies and violence and investigate whether motion-based anomaly discrimination reflects task-relevant motion information or observation-quality differences arising from person detection, pose estimation, and tracking. We introduce an observation-quality-aware evaluation framework that combines quality diagnostics, quality balancing, motion orthogonalization, and robustness analyses. On human-centered categories of UCF-Crime, quality-only features match or exceed the discrimination of multi-cue motion in the original cohort, revealing a strong observation-quality shortcut. After cardinality matching brings all 11 measured quality variables below a prespecified marginal balance threshold, multi-cue motion consistently outperforms quality-only features across Logistic Regression, Random Forest, and RBF-SVM, yielding ROC-AUC gains of 0.046-0.141. Because nonlinear quality-only prediction remains above chance after matching, we interpret this result as additional discrimination beyond measured marginal quality rather than as evidence of quality-free or causal motion understanding. These findings show that high VAD performance should not be interpreted as evidence of abnormal-motion modeling without auditing the reliability of the person-observation pipeline. Our study provides a systematic protocol for evaluating observation-quality shortcuts and for assessing whether motion-based discrimination survives explicit control of detection, pose-estimation, and tracking reliability.

---

### PATMT2: Two-stage domain adaptation for multilingual patent translation

Authors: Xiaowei Su, Xusen Cheng
Published: 2026-10-12
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2610.16466v1

We present PatMT2 (Patent Machine Translation, v2), a two-stage LoRA pipeline that adapts the Hy-MT2-7B dedicated translation model to four-language (Chinese/English/Japanese/Korean) patent translation. On six translation directions, the Stage-2 checkpoint reaches a macro-average BLEU of 43.39 (BF16) to 43.57 (AWQ INT4), improving over the Hy-MT2-7B base by up to +8.8 BLEU and also exceeding Qwen3-32B and several external systems on the same average. We then measure the cost of adaptation: probing 38 languages absent from the fine-tuning data, the model still produces correct, fluent output on 34, while four low-resource languages collapse into distinct, reproducible failure modes. We treat these as diagnostic cases rather than retention rates, and report their mechanisms without attributing them to a single cause. Paired bootstrap tests confirm the translation gains over all baselines (p < 0.001).

---

### Teaching Vision-Language Models to Decline: Support Completion Against Mirage

Authors: Alfaj Uddin Ahmed, Md. Shaown Miah, Md Arid Hasan, Syed Ishtiaque Ahmed, Sayeed Shafayet Chowdhury
Published: 2026-10-12
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.16465v1

Vision-language models answer questions about images they were never shown, a failure called mirage. Such an answer can be correct without being justified by the evidence, and the difference matters wherever a model's output is checked by someone who did not see its input. We locate the failure in the training objective: cross-entropy separates a model that reads the image from one that ignores it by only the conditional mutual information between answer and image, and no training example rewards declining. We formalise support completion, TIDE, which also shows each training question with a blank image, a same-domain wrong image and an out-of-domain image, with targets that decline, and prove that under TIDE every image-blind predictor is suboptimal by at least 1.04 nats per example, that the optimum declines exactly on the counterfactual kinds while keeping its behaviour on valid images, and that for any finite model the error of its answer-or-decline decision, declining on valid input plus answering on invalid input, is bounded below by how well its representation separates valid from invalid evidence. On Qwen2.5-VL-7B under a frozen held-out protocol with two judges, the three-condition predecessor of TIDE reduced the invalid-evidence mirage rate from 0.759 to 0.097 and the blank-image rate from 0.700 to 0.003 at a false-abstention cost of 0.126; label-free grounding raised the rate to 0.859, a likelihood-advantage contrast moved it only to 0.718 because its empty-image term carries no gradient, and grounding applied after support completion re-opened the mirage. The four-condition protocol on three base models, with seeds and component ablations, is in progress and not reported here.

---

### Online Individualized Treatment Selection under Partial Feedback: False Discovery Control and Minimax Optimality

Authors: Runzhe Gu, Zinan Zhao, Wenguang Sun, Xintao Xia
Published: 2026-10-12
Categories: stat.ML, cs.LG, stat.ME
arXiv: https://arxiv.org/abs/2610.16464v1

We study online individualized treatment selection when only the outcome under the assigned treatment is observed. Our objective is to preferentially treat individuals whose potential outcomes under treatment exceed a covariate-specific control reference, while limiting false discoveries among treated individuals. We propose Partial-Feedback Adaptive Individualized Treatment Selection (PAITS), which initializes learning from historical data and combines sequential estimation of local null risks, randomized exploration, and feedback-driven updates of the treatment selection threshold. Under regularity conditions, we establish asymptotic false discovery rate control and sublinear gap-weighted assignment regret relative to a population oracle. We further establish a minimax lower bound on assignment regret under prescribed uniform nuisance-estimation requirements and give conditions under which PAITS attains the corresponding constrained minimax rate up to logarithmic factors. These results quantify the tradeoff between accurate treatment assignment and nuisance learning under partial feedback. Simulation studies and a real-data analysis using the Jester recommendation dataset illustrate the finite-sample performance of PAITS.

---

### The interpreter as developer: An autoethnographic design study of building a computer-assisted interpreting tool through human-LLM co-creation

Authors: Zeming Zhao
Published: 2026-10-12
Categories: cs.CL
arXiv: https://arxiv.org/abs/2610.16460v1

Real-time AI captioning now serves audiences at multilingual events, yet computer-assisted interpreting (CAI) tools are still built for interpreters, rarely by them. This article reports an autoethnographic design study in which a professional interpreter with no programming training built a local-first CAI system through expert-led 'vibe coding', co-creating with an LLM architect and a coding agent. Analysis of the complete development dialogues shows how tacit professional knowledge, from segmentation to booth infrastructure, was externalized into design after generic defaults failed in use. In a demanding public scientific event, three mechanisms of in-booth appropriation emerge under a firmly human-led division of labor; telemetry from a first fully instrumented assignment shows a fully local chain sustaining a median translation latency of 0.94 s across 971 utterances. The case is read as lead-user innovation newly enabled by generative AI, with implications for the technological literacy of interpreters and the future of interpreter training.

---

### Differential Privacy Protected Preference Alignment on Survey Derived Data

Authors: Xiuyuan Qi, Yi Wang, Wenjie Wang
Published: 2026-10-12
Categories: cs.CR, cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2610.16458v1

Preference alignment enables language models to better reflect human values, but the preference data contains sensitive information, especially when derived from surveys about political beliefs, social norms, religion, gender attitudes, and cultural values. This paper studies privacy-preserving alignment on questionnaire-derived data from the World Values Survey, where supervision takes the form of ordinal labels rather than standard pairwise rankings. Existing efforts mainly focus on general privacy protections upon model training, while data-specific methods for survey data remain unexplored. We first show the vulnerability of aligned models, then formulate the task as rating prediction and introduce label differential privacy. Building on this formulation, we construct a differential private pipeline utilizing three privacy mechanisms: randomized response, exponential mechanism, and DP-SGD. Experiments show that simple label-level mechanisms outperform gradient-level methods, and proper aggregation function for preprocessing also make large contribution. The new fine-grained private comparison motivates explorations toward language model alignment with formal privacy guarantees.

---

### A temporal graph dissimilarity measure capturing epidemic spreading behavior

Authors: Salvatore Freni, Lorenzo Dall'Amico, Francesca Pistilli, Alain Barrat, Tatiana Tommasi, Ciro Cattuto
Published: 2026-10-12
Categories: cs.SI, cs.LG
arXiv: https://arxiv.org/abs/2610.16456v1

Comparing temporal networks is a challenging problem in network analysis. Commonly adopted methods compare structural properties of temporal graphs and can be highly sensitive to the pre-selected features. However, temporal networks are the support of dynamical processes, and a process may take place in statistically indistinguishable ways on structurally different temporal graphs. In particular, a dissimilarity measure that compares temporal graphs based on how spreading processes unfold on them is highly desirable. In this work, we first define the Time-Shift DTW, a dissimilarity between epidemic curves that accounts for both differences in their shape and temporal shifts. We then introduce Epidemic-Aware Temporal Graph Dissimilarity (EATGD), a deterministic and differentiable dissimilarity between temporal graphs that combines an embedding-based temporal graph distance with a set of topological and temporal features. Its parameters are calibrated using stochastic SIR simulations to maximize the correlation with the Time-Shift DTW, after which graph comparisons require no further epidemic simulations. Using empirical human proximity networks from the SocioPatterns collaboration and synthetic temporal graphs, we show that EATGD captures epidemic-based graph dissimilarity and generalizes across heterogeneous network settings. Its differentiability and computational efficiency make EATGD suitable for integration into machine-learning pipelines and provide a potential loss function for training generative models to reproduce epidemic dynamics without requiring an exact reconstruction of fine-grained network structure.

---

### GHOST: Geometry-Constrained Hybrid Diffusion for Open-Vocabulary 3D Scene Graph Generation

Authors: Changqun Feng, Wangxiandi Yin, Xin Hu, Ruiting Dai, Feiyang Yang, Jielei Wang, Tao He
Published: 2026-10-12
Categories: cs.CV, cs.MM
arXiv: https://arxiv.org/abs/2610.16455v1

3D scene graph generation requires understanding both object semantics and metric spatial relations. However, existing open-vocabulary methods often struggle with spatial predicates because language similarity reflects semantic co-occurrence rather than explicit geometry. We propose GHOST, a Geometry-constrained Hybrid diffusion framework for Open-vocabulary 3D Scene graph generaTion. GHOST jointly models continuous node geometry and discrete edge predicates, following a geometry-to-constraint-to-predicate denoising process to ensure geometrically consistent relation prediction. It introduces language-grounded constraint induction, which maps predicate embeddings to differentiable geometric constraints that guide both layout refinement and predicate reweighting. A Geometry-Topology Collaborative Denoiser further couples node and edge reasoning through noise-adaptive interaction and orientation-aware modeling. Experiments on 3DSSG demonstrate state-of-the-art performance in both closed-set and open-vocabulary settings, including over 17-point gains over RelationField on spatial predicates.

---

### Not Every Cue Is Evidence: Cue Supportability for Forensic Distillation in Face Forgery Detection

Authors: Binjia Zhou, Dawei Luo, Shuai Chen, Feng Xu, Haoyuan Li, Jia wen Seow, Jiachi Wang, Jie Lei, Zunlei Feng, Yijun Bei
Published: 2026-10-12
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.16452v1

Explainable face forgery detection requires not only correct judgments but also forensic evidence supported by the input image. However, training cues are typically derived from the forgery generation pipeline. Since images available to the model are of limited quality and resolution, it is difficult to observe the fine-grained traces these cues refer to. Training methods that further supply such cues as privileged supervision exacerbate the problem, as the student is taught to articulate evidence it cannot perceive. The student consequently resorts to templated or over-generalized expressions, giving rise to the correct-answer, wrong-evidence (CAWE) failure. We propose a dual supportability audit that filters privileged cues before distillation. Each candidate cue must be predictable from the remaining privileged context and causally tied to high-attribution visual regions in the current image. Only cues satisfying both conditions receive substantial distillation weight, preventing unsupported details from leaking into the student. Experiments across eight benchmarks show that our method maintains or improves detection accuracy while significantly raising explanation faithfulness compared to existing paradigms.

---

### StructZO: Structured Zeroth-Order Optimization for Memory-Efficient LLM Fine-Tuning

Authors: Xuanlin Yang, Jia Li, Zhouchen Lin
Published: 2026-10-12
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2610.16451v1

Fine-tuning large language models at scale is increasingly limited by the memory cost of backpropagation. Zeroth-Order (ZO) optimizers reduce this burden by relying only on forward function evaluations, but they still exhibit a substantial performance gap to first-order (FO) optimizers such as Adam. We argue that this gap does not arise solely from the absence of explicit gradients: FO fine-tuning also benefits from informative low-rank optimization geometry, large-scale signal aggregation, and well-conditioned update directions. Motivated by this, we introduce StructZO, a structured Zeroth-Order optimization framework for memory-efficient LLM fine-tuning. StructZO performs large-query ZO estimation within an activation-induced informative subspace and applies spectral control to the resulting reduced-space update before mapping it back to the original parameters. This lets ZO aggregate useful directional signals in a compact, geometry-aware space while avoiding the activation and optimizer-state memory costs of backpropagation. We show theoretically that, when the induced subspace captures concentrated optimization-relevant directions, StructZO reduces the effective ZO estimation dimension from $mn$ to $mr$, with the remaining projection error governed by the spectral tail of the activation Gram matrix. Empirically, across OPT-series and Llama3.1-8B on diverse downstreamtasks, StructZO improves over the strongest existing ZO baselines by up to $8.7%$ relative on average and up to $26.9%$ relative on individual tasks, while remaining within $1.5%$ of FO-Adam on average. These results suggest that structured and spectrally controlled ZO updates can substantially narrow the gap between ZO and FO optimization.

---

### CivWorldBench: Evaluating Language World Models in Structured Multi-Agent Sandbox Worlds

Authors: Ang Li, Haofei Yu, Zhengwei Tao, Liang Chen, Kuan Li, Miaomiao Li, Chenran Cai, Bin Liang, Jiaxuan You, Kam-Fai Wong
Published: 2026-10-12
Categories: cs.CL
arXiv: https://arxiv.org/abs/2610.16450v1

The growing use of language models for forecasting motivates a broader evaluation of their world-model capabilities. Such evaluation must go beyond predictive accuracy to assess hidden-state inference and intervention prediction. Real-world forecasting benchmarks offer limited access to hidden-state ground truth and controlled experiments. We introduce CivWorldBench, a benchmark built on a multi-agent strategy game engine with recorded internal states, controlled observations, and paired intervention rollouts. Its 1,900 decision points from 475 worlds support three task levels: association forecasting, belief inference, and intervention prediction. Reference predictors support attributable evaluation of capability gaps, with an expert harness and empirical ceiling estimates contextualizing scores. Stored trajectories make intermediate predictions verifiable, while agent-play experiments test whether world-model capabilities support decision-making. The main panel comprises eleven frontier language models. Composite benchmark scores correlate with real-world forecasting leaderboards on matched model panels, more closely than with a general-capability index, supporting its use as a proxy for real-world forecasting capability. Strong association forecasts do not consistently extend to belief inference or intervention prediction, and effect magnitudes remain poorly calibrated despite informative direction predictions. Process verification reveals errors missed by final-answer evaluation, and structured reasoning improves intervention prediction for the strongest model in the process study. Agent-play experiments identify useful assisted plans and motivate evaluating decision-relevant capabilities beyond association forecasting. These findings identify hidden-state inference, intermediate-state prediction, and effect calibration as priorities for world-model training.

---

### CryptoFlowBench: Benchmarking Multi-Hop Reasoning on Blockchains for Cryptocurrency Fund Tracing

Authors: Runang He, Tongya Zheng, Jing Cao, Yucheng Ying, Shunyu Liu, Bingde Hu, Yang Gao, Zeyuan Hu, Jianrong Zhang, Mingli Song
Published: 2026-10-12
Categories: cs.CR, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2610.16449v1

Pseudonymous accounts and complex smart-contract ecosystems substantially hinder the tracing of illicit cryptocurrency fund flows. Existing Graph Neural Networks (GNNs) have achieved promising performance in blockchain fraud detection, yet their static node- or edge-level predictions do not naturally generalize to multi-hop fund tracing. Meanwhile, existing tracing methods mainly rely on predefined heuristics to capture multi-hop relationships, limiting their adaptive reasoning capabilities. To bridge this gap, we introduce CryptoFlowBench, a benchmark constructed from authoritative reports on cryptocurrency-related illicit incidents, with four tasks spanning local to multi-hop fund tracing. These tasks pose substantial challenges to existing learning- and heuristic-based methods, particularly in complex multi-hop scenarios. We therefore propose ChainSleuth, a graph-tool-augmented agent that iteratively reasons, invokes graph tools, and updates its tracing trajectory, enabling adaptive information acquisition and recovery from erroneous paths. Experiments show that no baseline consistently dominates all four tasks, revealing substantial gaps between local fraud detection and multi-hop fund tracing. ChainSleuth achieves notable improvements on multi-hop tasks, demonstrating the potential of agentic graph reasoning for cryptocurrency fund investigation.

---

### Tikhonov-Stabilized Bezier Representation Forecasting for Training-free Diffusion Acceleration

Authors: Lei Zhu, Mujie Lin, Ruochong Zheng, Guangyi Wang, Hao Li, Peng Jin, Chang Liu, Jie Chen
Published: 2026-10-12
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.16448v1

Diffusion models, particularly Diffusion Transformers, achieve strong image and video generation quality but remain expensive at inference time due to repeated denoiser evaluations. Feature caching offers a deployment-friendly acceleration strategy by reusing or predicting intermediate representations without retraining the generator or modifying the sampler. However, existing cache-then-forecast methods often rely on Taylor-style finite-difference extrapolation, which can become unstable over longer cache intervals, and typically predict local module outputs whose errors may accumulate through subsequent denoiser blocks. We propose \textbf{BeziCast}, a training-free diffusion acceleration framework that forecasts output-proximal denoising representations using low-order B'ezier trajectories. BeziCast estimates B'ezier control points via Tikhonov-stabilized fitting, providing a smooth, capacity-controlled temporal parameterization that avoids explicit high-order derivative estimation and decouples trajectory capacity from cache-query density. To moderate aggressive extrapolation, we further derive equivalent forecasting weights and introduce a convex-hull-inspired guardrail that detects high-risk predictions and softly pulls them toward a conservative simplex-projected estimate. Extensive evaluations across advanced image and video diffusion models demonstrate the effectiveness of BeziCast. In particular, BeziCast accelerates FLUX.1 by up to (4.79\times) and HunyuanVideo by up to (4.11\times), while preserving substantially better generation quality than competing acceleration baselines.

---

### Explaining Heterogeneous Treatment Effects with Causal Forests and the Hoeffding Functional Decomposition

Authors: Gaëtan Patinier, Clément Bénard, Margaux Törnqvist
Published: 2026-10-12
Categories: stat.ML, cs.LG
arXiv: https://arxiv.org/abs/2610.16447v1

The estimation of heterogeneous treatment effects is a central objective of causal inference. However, state-of-the-art algorithms to solve this task are based on machine learning principles, and thus behave as black boxes. Consequently, the identification of covariates driving heterogeneity remains a challenging and open problem. To address this limitation, we introduce CausalTreeHFD, an interpretable extension of causal forests, based on the Hoeffding functional decomposition adapted to tree ensembles. Our approach decomposes predictions of conditional average treatment effects (CATE) into contributions from individual covariates and pairwise interactions, providing a direct characterization of the factors driving treatment effect heterogeneity. We establish the consistency of CausalTreeHFD, which converges towards the Hoeffding decomposition of the underlying CATE function, under standard assumptions for the analysis of causal forests. Experiments on simulated and real datasets show that CausalTreeHFD accurately identifies relevant covariates and interactions, and precisely recovers their contributions to heterogeneous treatment effects.

---

### Pixels or Code? Benchmarking Instruction-Guided Visual Artifact Operation

Authors: Shun Guo, Shuizhou Shi, Boao Wang, Wei Xiong, Siyu Sun, Jiangtong Li, Jie Li
Published: 2026-10-12
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.16444v1

Visual artifacts such as charts, technical diagrams, webpages, and document pages underpin scientific, commercial, and engineering workflows. Editing these artifacts requires direct modification, dependency-aware updates, preservation of unaffected content, and compliance with domain constraints. We define this capability as Visual Artifact Operation (VAO). Prior benchmarks test parts of this problem, including cascading updates in charts. We study direct edits, dependent updates, and preservation across visual domains and pixel, code, and agent paradigms under one image-only protocol. We introduce Visual Artifact Operation Bench (VAO-Bench), comprising 1,573 manually screened tasks across nine artifact categories and three difficulty levels. VAO-Bench uses executable artifacts and deterministic rendering to trace reference transformations and annotate direct edits, dependent updates, and preservation constraints. All systems receive the same source image and instruction; we evaluate pixel-space editors, code-generating multimodal systems, and tool-using agents on direct-edit accuracy, dependency-aware update, content preservation, domain validity, and visual quality. Experiments show that executable-code systems better support dependent updates, while pixel-space editors offer faster execution and strong visual quality. Explicitly stating dependencies improves both paradigms, identifying implicit-update inference as a central bottleneck. VAO-Bench provides a reproducible foundation for dependable visual artifact editing across generation paradigms.

---

### RAP-WAM: Learning World-Action Models through Predictive Rehearsal and Action Perturbations

Authors: Yuanpeng He, Ceyao Zhang, Fangjing Li, Kefei Zhu, Lijian Li, Yueheng Li, Tianjia He, Wenpin Jiao, Zhi Jin, Yaodong Yang, Yuanpei Chen
Published: 2026-10-12
Categories: cs.AI
arXiv: https://arxiv.org/abs/2610.16443v2

Recent world action models draw on video generation priors to jointly predict future observations and robot actions. Adapting these priors to control requires determining how future predictions should guide policy learning. RAP-WAM combines predictive Rehearsal and Action Perturbations (RAP) with future-supervised queries and Coupled Action Projection for policy learning. Predictive rehearsal trains on model-generated inputs while retaining recorded targets, and accumulates recurring errors into a spatial prior. We perturb actions and measure changes in video predictions. Action-response-based spatiotemporal credit assignment combines these responses, prediction error, and the rehearsal prior to allocate visual supervision; spatially aggregating the responses determines supervision intensity across future action steps. No separate scoring network or deployment-time probe is required. The base learns from a video-pretrained backbone without intermediate pretraining on external embodied data. We evaluate the fixed base and its agent-refined, task-conditioned system. On 42 RoboDojo tasks, their average success rates are 12.85% and 15.62%, with Scores of 17.86 and 20.59. Across 15 real-robot tasks, the complete system achieves 89.00% task-macro-average success, exceeding the strongest of five baselines by 37.67 percentage points. Across 315 fixed-output simulation windows, 45.21% of response-top20 positions fall outside error-top20, revealing information beyond prediction error. Videos and further details are available on the project website: https://rapwam.github.io/.

---

### VisExMEM: Visual Evidence-Based Explainable Multimodal Evaluation Metric

Authors: Subhadeep Roy, Christoph Leiter, Steffen Eger
Published: 2026-10-12
Categories: cs.CV, cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2610.16441v1

Current image-text evaluation often relies on a global score, which gives limited insight into which semantic details are supported by the image. Explainable metrics provide rationales, but the visual evidence supporting those rationales can still be difficult to verify. We introduce VisExMEM, a reference-free evaluator that decomposes captions into atomic claims, localizes visual evidence, verifies each claim, and averages claim-level support into a global score. Explanations use the same evidence and decisions. We train Qwen3.5-based 4B and 9B models on 100K samples with multi-task SFT, counterfactual refinement, and evidence-aware GDPO. Across eight benchmarks, VisExMEM-9B achieves the highest descriptive composite among open-source evaluators in our comparison, exceeding VQAScore by 11.7 points while remaining competitive with GPT-5.6 Sol. We also introduce AGITA-Bench, containing 600 human-annotated image-text pairs and 8.7K atomic claims for global and claim-level evaluation. For untuned Qwen3.5 and GPT-5.6 Sol controls, decomposed prompting raises the same composite by 3.6-5.1 points relative to holistic prompting. Human evaluation shows that VisExMEM explanations outperform prior explainable metrics and approach GPT-5.6 Sol.

---

### ARMED: Adaptive Retrieval-Augmented Medication Recommendation

Authors: Kyungkeon Chung, Se Un Park, Hyejin Lee, Junyoung Park
Published: 2026-10-12
Categories: cs.LG, cs.AI, cs.CL, cs.IR
arXiv: https://arxiv.org/abs/2610.16440v1

Medication recommendation requires jointly considering diagnoses, comorbidities, longitudinal patient history, and treatment interactions. Although recent retrieval-augmented generation approaches using language models have shown promise for clinical recommendation, most existing systems rely on fixed retrieval pipelines that cannot adapt to diverse clinical retrieval scenarios and may introduce noisy or irrelevant evidence. We propose ARMED, an adaptive clinical RAG framework that dynamically expands retrieval according to retrieval sufficiency and current-admission relevance. ARMED first observes the initial retrieval outcome and selectively performs supplementary or fallback multi-query retrieval when additional evidence is needed. The framework also incorporates a relevance gate that activates only clinically relevant same-patient history and organizes retrieved evidence into structured query-level clinical evidence. Experiments on MIMIC-IV demonstrate that ARMED consistently outperforms previous static retrieval baselines and improves overall medication recommendation quality. By combining locally runnable small language models with adaptive retrieval, ARMED supports medication recommendation without relying on external language model APIs or task-specific fine-tuning.

---

### CultureBot: A Framework for Retrieval-Augmented Exploration of Cultural Heritage Data

Authors: Vasileios Karampinis, Angelos Vlachos, Giorgos Filandrianos, Maria Lymperaiou, Paraskevi Tzouveli, Chrysoula Zerva, Giorgos Stamou
Published: 2026-10-12
Categories: cs.CL
arXiv: https://arxiv.org/abs/2610.16435v1

Digital cultural heritage collections encode rich relations among objects, places, periods, providers, images, and source institutions, yet users typically access them through keyword search and faceted browsing. We present CULTUREBOT, an interactive framework for natural-language exploration of cultural heritage collections that combines knowledge-graph-guided retrieval, lexical-semantic retrieval, and source-linked generation. CULTUREBOT maps user queries to structured catalogue relations, enriches retrieved records with textual evidence, and generates answers that link back to the underlying repository records. We instantiate the framework on 81,852 SearchCulture records and release an online demonstration. In an expert evaluation involving six cultural-heritage specialists, 55 persona-grounded queries, and 250 evaluations, the graph-guided system was rated significantly higher than two general-purpose assistants and a non-graph ablation, with the largest gains in evidence grounding, diversity, reasoning, and completeness. These results support graph-guided retrieval as a practical way to expose the relational structure of cultural collections through conversational access while preserving links to catalogue evidence.

---

### ECHO: Correcting Time Series Foundation Model Forecasts with Historical Analogues and Errors

Authors: Boya Zhang, Xilin Dai, Yilin Wang, Qiang Xu, Luyao Dong, Xing He, Wen Wu
Published: 2026-10-12
Categories: cs.LG
arXiv: https://arxiv.org/abs/2610.16433v1

Time Series Foundation Models (TSFMs) provide general-purpose forecasting capabilities, yet their forecasts can remain misaligned with the temporal characteristics of a target domain. This motivates lightweight forecast correction even when only black-box access to the TSFM is available. Our key observation is that two forms of historical information provide complementary correction signals: historical analogues reveal how similar contexts evolved, while past forecast errors reveal a TSFM's characteristic deviations in the target domain. Based on this observation, we propose ECHO, a black-box forecast correction framework that learns from both sources. ECHO uses AnalogE to assess retrieved historical analogues together with the current context and base forecast, learning transferable corrections through cross-domain pretraining. In parallel, RecentE learns target-specific residual corrections from past forecast errors and refines them as causally available feedback accumulates. ECHO dynamically combines the two corrections according to their observed forecasting performance. Across 5 representative TSFMs and 9 real-world datasets, ECHO reduces MSE in all 45 backbone--dataset settings, yielding an average relative reduction of 9.26\% in macro-average MSE. Its gains also extend to long-horizon forecasting.

---

### Taming Nonconvexity with Epigraph Lifting

Authors: Clara Lage, Hok Shing Wong, Matthias Ehrhardt
Published: 2026-10-12
Categories: math.OC, cs.LG
arXiv: https://arxiv.org/abs/2610.16429v1

We consider inverse problems regularized by a trained neural network with \emph{convex activations} and \emph{unrestricted weights}. The resulting optimization problem is generally nonconvex, and gradient-based methods may converge to poor local minima. We introduce a \emph{penalized epigraph lifting} of the network and show that the lifted problem admits a \emph{difference-of-convex (DC) decomposition}. For Lipschitz continuous activations, we establish \emph{finite exactness}: the original objective is recovered exactly at a finite and explicitly computable penalty threshold $\barρ$. Building on this structure, we propose \emph{H-DC} (Homotopic DC), a homotopic scheme that solves a sequence of DC subproblems with increasing $ρ$, warm-starting each level from the previous solution. Numerical experiments on synthetic problems, a manifold denoiser regularizer, and image inpainting with a learned Mumford--Shah regularizer show that H-DC consistently escapes poor local minima and reaches solutions with lower objective values than gradient-based methods from the same initialization.

---

### Resample the Answer: Where to Spend Test-Time Compute in Recursive Reasoning

Authors: Maxime Illouz
Published: 2026-10-12
Categories: cs.LG
arXiv: https://arxiv.org/abs/2610.16427v1

Recursive reasoning models such as the Tiny Recursive Model (TRM) iteratively refine two latent states: an answer state y, decoded into the output, and a reasoning state z, which is never decoded. Many of their failures have been attributed to bad basins, regions of latent space that trap the recursion on wrong answers. Recent test-time approaches perturb y, z or both to recover these failures, but the impact of this choice on performance, test-time compute efficiency and scaling properties remains unclear. Through controlled experiments on TRM, we find that a failed trajectory carries its failure in its answer state. At equal compute, in depth and in width, resampling y recovers more failures than any perturbation of z we tested, and the performance gap widens as the test-time compute budget grows. Our approach, Gated Answer Resampling (GAR), pairs this resampling of the answer state with the model's own halting head and restarts only the puzzles that the head rejects. With at most 64 restarts per puzzle and no training or hyperparameter search, GAR takes TRM from 85.9% to 99.4% on Sudoku-Extreme and from 68.8% to 85.6% on Pencil Puzzle Bench validation. On the same checkpoints, GAR also outperforms Probabilistic TRM (PTRM), which perturbs z at each step, even when PTRM uses 42 to 197 times as much compute.

---

### AuthGateBench: Benchmarking LLMs as Authorization Gate

Authors: Zhixiong Zhuang, Jin Jia, Zhiwei Shang, Ziyu Wang
Published: 2026-10-12
Categories: cs.AI
arXiv: https://arxiv.org/abs/2610.16426v1

Automated approval systems let coding agents execute high-risk actions with minimal human oversight by delegating the authorization decision to a classifier, typically a large language model (LLM). Such systems, adopted in tools like Claude Code and Codex, remain vulnerable to failure modes such as scope escalation, where the agent takes actions exceeding what the user authorized. Existing benchmarks focus mainly on malicious requests, evaluating refusal under an implicit policy. We introduce AuthGateBench, the first policy-guided benchmark for authorization gating under benign requests. We implement a novel pipeline where an LLM generates contrast pairs with identical action and trajectory but differing request, flipping the label between block and allow. This allows us to assess whether models reason over context rather than the action alone. We further examine whether auxiliary environment signals can help models make better decisions. The resulting benchmark comprises 1,400 instances spanning five failure modes drawn from real-world Claude Code incidents. Evaluations show accuracy degrades unevenly across failure modes as models over-rely on internal safety alignment, and models still block most actions even when explicitly authorized, revealing unstable rather than deterministic decisions; retrospectively retrieved environment signals, however, substantially improve authorization accuracy.

---