# Research Papers — 2026-01 · bundle page 120 of 189

> Full-text markdown of papers #5951–6000 of 9403 published in 2026-01, newest first. AI-readable mirrors of the full 9403-paper 2026-01 archive (50 papers per file).

Canonical page: https://aipod.dev/papers/archive/2026-01/page/120
[← Previous bundle](https://aipod.dev/papers/archive/2026-01/page/119.md)
[Next bundle →](https://aipod.dev/papers/archive/2026-01/page/121.md)

### Rewarding the Rare: Uniqueness-Aware RL for Creative Problem Solving in LLMs

Authors: Zhiyuan Hu, Yucheng Wang, Yufei He, Jiaying Wu, Yilun Zhao, See-Kiong Ng, Cynthia Breazeal, Anh Tuan Luu, Hae Won Park, Bryan Hooi
Published: 2026-01-13
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2601.08763v2

Reinforcement learning (RL) has become a central paradigm for post-training large language models (LLMs), particularly for complex reasoning tasks, yet it often suffers from exploration collapse: policies prematurely concentrate on a small set of dominant reasoning patterns, improving pass@1 while limiting rollout-level diversity and gains in pass@k. We argue that this failure stems from regularizing local token behavior rather than diversity over sets of solutions. To address this, we propose Uniqueness-Aware Reinforcement Learning, a rollout-level objective that explicitly rewards correct solutions that exhibit rare high-level strategies. Our method uses an LLM-based judge to cluster rollouts for the same problem according to their high-level solution strategies, ignoring superficial variations, and reweights policy advantages inversely with cluster size. As a result, correct but novel strategies receive higher rewards than redundant ones. Across mathematics, physics, and medical reasoning benchmarks, our approach consistently improves pass@$k$ across large sampling budgets and increases the area under the pass@$k$ curve (AUC@$K$) without sacrificing pass@1, while sustaining exploration and uncovering more diverse solution strategies at scale.

---

### Adaptive Requesting in Decentralized Edge Networks via Non-Stationary Bandits

Authors: Yi Zhuang, Kun Yang, Xingran Chen
Published: 2026-01-13
Categories: cs.LG, cs.MA
arXiv: https://arxiv.org/abs/2601.08760v3

We study a decentralized collaborative requesting problem that aims to optimize the information freshness of time-sensitive clients in edge networks consisting of multiple clients, access nodes (ANs), and servers. Clients request content through ANs acting as gateways, without observing AN states or the actions of other clients. We define the reward as the age of information reduction resulting from a client's selection of an AN, and formulate the problem as a non-stationary multi-armed bandit. In this decentralized and partially observable setting, the resulting reward process is history-dependent and coupled across clients, and exhibits both abrupt and gradual changes in expected rewards, rendering classical bandit-based approaches ineffective. To address these challenges, we propose the AGING BANDIT WITH ADAPTIVE RESET algorithm, which combines adaptive windowing with periodic monitoring to track evolving reward distributions. We establish theoretical performance guarantees showing that the proposed algorithm achieves near-optimal performance, and we validate the theoretical results through simulations.

---

### M3CoTBench: Benchmark Chain-of-Thought of MLLMs in Medical Image Understanding

Authors: Juntao Jiang, Jiangning Zhang, Yali Bi, Jinsheng Bai, Weixuan Liu, Weiwei Jin, Zhucun Xue, Yong Liu, Xiaobin Hu, Shuicheng Yan
Published: 2026-01-13
Categories: eess.IV, cs.CV
arXiv: https://arxiv.org/abs/2601.08758v4

Chain-of-Thought (CoT) reasoning has proven effective in enhancing large language models by encouraging step-by-step intermediate reasoning, and recent advances have extended this paradigm to Multimodal Large Language Models (MLLMs). In the medical domain, where diagnostic decisions depend on nuanced visual cues and sequential reasoning, CoT aligns naturally with clinical thinking processes. However, current benchmarks for medical image understanding generally focus on the final answer while ignoring the reasoning path. Such opaque reasoning processes lack reliable bases for judgment, making it difficult to assist doctors in diagnosis. To address this gap, we introduce a new M3CoTBench benchmark specifically designed to evaluate the correctness, efficiency, impact, and consistency of CoT reasoning in medical image understanding. M3CoTBench features 1) a diverse, multi-level difficulty dataset covering 24 examination types, 2) 13 varying-difficulty tasks, 3) a suite of CoT-specific evaluation metrics (correctness, efficiency, impact, and consistency) tailored to clinical reasoning, and 4) a performance analysis of multiple MLLMs. M3CoTBench systematically evaluates CoT reasoning across diverse medical imaging tasks, revealing current limitations of MLLMs in generating reliable and clinically interpretable reasoning, and aims to foster the development of transparent, trustworthy, and diagnostically accurate AI systems for healthcare. Project page at https://juntaojianggavin.github.io/projects/M3CoTBench/.

---

### Grid-Aware Charging and Operational Optimization for Mixed-Fleet Public Transit

Authors: Rishav Sen, Amutheezan Sivagnanam, Aron Laszka, Ayan Mukhopadhyay, Abhishek Dubey
Published: 2026-01-13
Categories: math.OC, cs.AI, eess.SY
arXiv: https://arxiv.org/abs/2601.08753v1

The rapid growth of urban populations and the increasing need for sustainable transportation solutions have prompted a shift towards electric buses in public transit systems. However, the effective management of mixed fleets consisting of both electric and diesel buses poses significant operational challenges. One major challenge is coping with dynamic electricity pricing, where charging costs vary throughout the day. Transit agencies must optimize charging assignments in response to such dynamism while accounting for secondary considerations such as seating constraints. This paper presents a comprehensive mixed-integer linear programming (MILP) model to address these challenges by jointly optimizing charging schedules and trip assignments for mixed (electric and diesel bus) fleets while considering factors such as dynamic electricity pricing, vehicle capacity, and route constraints. We address the potential computational intractability of the MILP formulation, which can arise even with relatively small fleets, by employing a hierarchical approach tailored to the fleet composition. By using real-world data from the city of Chattanooga, Tennessee, USA, we show that our approach can result in significant savings in the operating costs of the mixed transit fleets.

---

### A Geolocation-Aware Multimodal Approach for Ecological Prediction

Authors: Valerie Zermatten, Chiara Vanalli, Gencer Sumbul, Diego Marcos, Devis Tuia
Published: 2026-01-13
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.08750v2

While integrating multiple modalities has the potential to improve environmental monitoring, current approaches struggle to combine data sources with heterogeneous formats or contents. A central difficulty arises when combining continuous gridded data (e.g., remote sensing) with sparse and irregular point observations such as species records. Existing geostatistical and deep-learning-based approaches typically operate on a single modality or focus on spatially aligned inputs, and thus cannot seamlessly overcome this difficulty. We propose a Geolocation-Aware MultiModal Approach (GAMMA), a transformer-based fusion approach designed to integrate heterogeneous ecological data using explicit spatial context. Instead of interpolating observations into a common grid, GAMMA first represents all inputs as location-aware embeddings that preserve spatial relationships between samples. GAMMA dynamically selects relevant neighbours across modalities and spatial scales, enabling the model to jointly exploit continuous remote sensing imagery and sparse geolocated observations. We evaluate GAMMA on the task of predicting 103 environmental variables from the SWECO25 data cube across Switzerland. Inputs combine aerial imagery with biodiversity observations from GBIF and textual habitat descriptions from Wikipedia, provided by the EcoWikiRS dataset. Experiments show that multimodal fusion consistently improves prediction performance over single-modality baselines and that explicit spatial context further enhances model accuracy. The flexible architecture of GAMMA also allows to analyse the contribution of each modality through controlled ablation experiments. These results demonstrate the potential of location-aware multimodal learning for integrating heterogeneous ecological data and for supporting large-scale environmental mapping tasks and biodiversity monitoring.

---

### A Single-Parameter Factor-Graph Image Prior

Authors: Tianyang Wang, Ender Konukoglu, Hans-Andrea Loeliger
Published: 2026-01-13
Categories: eess.IV, cs.CV, eess.SP
arXiv: https://arxiv.org/abs/2601.08749v2

We propose a novel piecewise smooth image model with piecewise constant local parameters that are automatically adapted to each image. Technically, the model is formulated in terms of factor graphs with NUP (normal with unknown parameters) priors, and the pertinent computations amount to iterations of conjugate-gradient steps and Gaussian message passing. The proposed model and algorithms are demonstrated with applications to denoising and contrast enhancement.

---

### To Retrieve or To Think? Cross-Boundary Context Evolution for Multi-hop Complex Reasoning

Authors: Rubing Chen, Jian Wang, Wenjie Li, Xiao-Yong Wei, Qing Li
Published: 2026-01-13
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2601.08747v3

Current context augmentation methods, such as retrieval-augmented generation, play a crucial role in bridging a model's internal knowledge boundary and external evidence for multi-hop reasoning. However, they often follow a rigid policy and treat external retrieval as the default action at each step. Such brute-force context expansion incurs unnecessary computational cost and may degrade reasoning performance by saturating the context with redundant or weakly relevant evidence. In this paper, we propose cross-boundary Context Evolution (EvoCtx), a framework that models complex reasoning as an adaptive process of boundary-aware context evolution. EvoCtx dynamically decides whether the next reasoning transition should cross the current evidence boundary through retrieval or refine the reasoning state within the existing context. It estimates the semantic gap between the reasoning state and the accumulated evidence, and strategically alternates between boundary expansion and intra-boundary trajectory refinement. This eliminates redundant retrieval steps and preserves a compact, evidence-supported reasoning trajectory. Extensive experiments on challenging open-domain and multi-hop QA benchmarks demonstrate that EvoCtx significantly outperforms previous methods, offering an effective approach to complex reasoning tasks. The source code can be accessed at https://github.com/Anya-RB-Chen/EvoCtx.

---

### TableCache: Primary Foreign Key Guided KV Cache Precomputation for Low Latency Text-to-SQL

Authors: Jinbo Su, Yuxuan Hu, Cuiping Li, Hong Chen, Jia Li, Lintao Ma, Jing Zhang
Published: 2026-01-13
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2601.08743v1

In Text-to-SQL tasks, existing LLM-based methods often include extensive database schemas in prompts, leading to long context lengths and increased prefilling latency. While user queries typically focus on recurrent table sets-offering an opportunity for KV cache sharing across queries-current inference engines, such as SGLang and vLLM, generate redundant prefix cache copies when processing user queries with varying table orders. To address this inefficiency, we propose precomputing table representations as KV caches offline and querying the required ones online. A key aspect of our approach is the computation of table caches while preserving primary foreign key relationships between tables. Additionally, we construct a Table Trie structure to facilitate efficient KV cache lookups during inference. To enhance cache performance, we introduce a cache management system with a query reranking strategy to improve cache hit rates and a computation loading pipeline for parallelizing model inference and cache loading. Experimental results show that our proposed TableCache achieves up to a 3.62x speedup in Time to First Token (TTFT) with negligible performance degradation.

---

### Inferring Latent Intentions: Attributional Natural Language Inference in LLM Agents

Authors: Xin Quan, Jiafeng Xiong, Marco Valentino, André Freitas
Published: 2026-01-13
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.08742v1

Attributional inference, the ability to predict latent intentions behind observed actions, is a critical yet underexplored capability for large language models (LLMs) operating in multi-agent environments. Traditional natural language inference (NLI), in fact, fails to capture the nuanced, intention-driven reasoning essential for complex interactive systems. To address this gap, we introduce Attributional NLI (Att-NLI), a framework that extends NLI with principles from social psychology to assess an agent's capacity for abductive intentional inference (generating hypotheses about latent intentions), and subsequent deductive verification (drawing valid logical conclusions). We instantiate Att-NLI via a textual game, Undercover-V, experimenting with three types of LLM agents with varying reasoning capabilities and access to external tools: a standard NLI agent using only deductive inference, an Att-NLI agent employing abductive-deductive inference, and a neuro-symbolic Att-NLI agent performing abductive-deductive inference with external theorem provers. Extensive experiments demonstrate a clear hierarchy of attributional inference capabilities, with neuro-symbolic agents consistently outperforming others, achieving an average win rate of 17.08%. Our results underscore the role that Att-NLI can play in developing agents with sophisticated reasoning capabilities, highlighting, at the same time, the potential impact of neuro-symbolic AI in building rational LLM agents acting in multi-agent environments.

---

### From Rows to Reasoning: A Retrieval-Augmented Multimodal Framework for Spreadsheet Understanding

Authors: Anmol Gulati, Sahil Sen, Waqar Sarguroh, Kevin Paul
Published: 2026-01-13
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.08741v2

Large Language Models (LLMs) struggle to reason over large-scale enterprise spreadsheets containing thousands of numeric rows, multiple linked sheets, and embedded visual content such as charts and receipts. Prior state-of-the-art spreadsheet reasoning approaches typically rely on single-sheet compression or full-context encoding, which limits scalability and fails to reflect how real users interact with complex, multimodal workbooks. We introduce FRTR-Bench, the first large-scale benchmark for multimodal spreadsheet reasoning, comprising 30 enterprise-grade Excel workbooks spanning nearly four million cells and more than 50 embedded images. To address these challenges, we present From Rows to Reasoning (FRTR), an advanced, multimodal retrieval-augmented generation framework that decomposes Excel workbooks into granular row, column, and block embeddings, employs hybrid lexical-dense retrieval with Reciprocal Rank Fusion (RRF), and integrates multimodal embeddings to reason over both numerical and visual information. We tested FRTR on six LLMs, achieving 74% answer accuracy on FRTR-Bench with Claude Sonnet 4.5, a substantial improvement over prior state-of-the-art approaches that reached only 24%. On the SpreadsheetLLM benchmark, FRTR achieved 87% accuracy with GPT-5 while reducing token usage by roughly 50% compared to direct serialization methods.

---

### PrivGemo: Privacy-Preserving Dual-Tower Graph Retrieval for Empowering LLM Reasoning with Memory Augmentation

Authors: Xingyu Tan, Xiaoyang Wang, Qing Liu, Xiwei Xu, Xin Yuan, Liming Zhu, Wenjie Zhang
Published: 2026-01-13
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.08739v1

Knowledge graphs (KGs) provide structured evidence that can ground large language model (LLM) reasoning for knowledge-intensive question answering. However, many practical KGs are private, and sending retrieved triples or exploration traces to closed-source LLM APIs introduces leakage risk. Existing privacy treatments focus on masking entity names, but they still face four limitations: structural leakage under semantic masking, uncontrollable remote interaction, fragile multi-hop and multi-entity reasoning, and limited experience reuse for stability and efficiency. To address these issues, we propose PrivGemo, a privacy-preserving retrieval-augmented framework for KG-grounded reasoning with memory-guided exposure control. PrivGemo uses a dual-tower design to keep raw KG knowledge local while enabling remote reasoning over an anonymized view that goes beyond name masking to limit both semantic and structural exposure. PrivGemo supports multi-hop, multi-entity reasoning by retrieving anonymized long-hop paths that connect all topic entities, while keeping grounding and verification on the local KG. A hierarchical controller and a privacy-aware experience memory further reduce unnecessary exploration and remote interactions. Comprehensive experiments on six benchmarks show that PrivGemo achieves overall state-of-the-art results, outperforming the strongest baseline by up to 17.1%. Furthermore, PrivGemo enables smaller models (e.g., Qwen3-4B) to achieve reasoning performance comparable to that of GPT-4-Turbo.

---

### TerraFormer: Automated Infrastructure-as-Code with LLMs Fine-Tuned via Policy-Guided Verifier Feedback

Authors: Prithwish Jana, Sam Davidson, Bhavana Bhasker, Andrey Kan, Anoop Deoras, Laurent Callot
Published: 2026-01-13
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2601.08734v1

Automating Infrastructure-as-Code (IaC) is challenging, and large language models (LLMs) often produce incorrect configurations from natural language (NL). We present TerraFormer, a neuro-symbolic framework for IaC generation and mutation that combines supervised fine-tuning with verifier-guided reinforcement learning, using formal verification tools to provide feedback on syntax, deployability, and policy compliance. We curate two large, high-quality NL-to-IaC datasets, TF-Gen (152k instances) and TF-Mutn (52k instances), via multi-stage verification and iterative LLM self-correction. Evaluations against 17 state-of-the-art LLMs, including ~50x larger models like Sonnet 3.7, DeepSeek-R1, and GPT-4.1, show that TerraFormer improves correctness over its base LLM by 15.94% on IaC-Eval, 11.65% on TF-Gen (Test), and 19.60% on TF-Mutn (Test). It outperforms larger models on both TF-Gen (Test) and TF-Mutn (Test), ranks third on IaC-Eval, and achieves top best-practices and security compliance.

---

### A Novel Approach to Explainable AI with Quantized Active Ingredients in Decision Making

Authors: A. M. A. S. D. Alagiyawanna, Asoka Karunananda, Thushari Silva, A. Mahasinghe
Published: 2026-01-13
Categories: cs.LG, quant-ph
arXiv: https://arxiv.org/abs/2601.08733v1

Artificial Intelligence (AI) systems have shown good success at classifying. However, the lack of explainability is a true and significant challenge, especially in high-stakes domains, such as health and finance, where understanding is paramount. We propose a new solution to this challenge: an explainable AI framework based on our comparative study with Quantum Boltzmann Machines (QBMs) and Classical Boltzmann Machines (CBMs). We leverage principles of quantum computing within classical machine learning to provide substantive transparency around decision-making. The design involves training both models on a binarised and dimensionally reduced MNIST dataset, where Principal Component Analysis (PCA) is applied for preprocessing. For interpretability, we employ gradient-based saliency maps in QBMs and SHAP (SHapley Additive exPlanations) in CBMs to evaluate feature attributions.QBMs deploy hybrid quantum-classical circuits with strongly entangling layers, allowing for richer latent representations, whereas CBMs serve as a classical baseline that utilises contrastive divergence. Along the way, we found that QBMs outperformed CBMs on classification accuracy (83.5% vs. 54%) and had more concentrated distributions in feature attributions as quantified by entropy (1.27 vs. 1.39). In other words, QBMs not only produced better predictive performance than CBMs, but they also provided clearer identification of "active ingredient" or the most important features behind model predictions. To conclude, our results illustrate that quantum-classical hybrid models can display improvements in both accuracy and interpretability, which leads us toward more trustworthy and explainable AI systems.

---

### ISLA: A U-Net for MRI-based acute ischemic stroke lesion segmentation with deep supervision, attention, domain adaptation, and ensemble learning

Authors: Vincent Roca, Martin Bretzner, Hilde Henon, Laurent Puy, Grégory Kuchcinski, Renaud Lopes
Published: 2026-01-13
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2601.08732v1

Accurate delineation of acute ischemic stroke lesions in MRI is a key component of stroke diagnosis and management. In recent years, deep learning models have been successfully applied to the automatic segmentation of such lesions. While most proposed architectures are based on the U-Net framework, they primarily differ in their choice of loss functions and in the use of deep supervision, residual connections, and attention mechanisms. Moreover, many implementations are not publicly available, and the optimal configuration for acute ischemic stroke (AIS) lesion segmentation remains unclear. In this work, we introduce ISLA (Ischemic Stroke Lesion Analyzer), a new deep learning model for AIS lesion segmentation from diffusion MRI, trained on three multicenter databases totaling more than 1500 AIS participants. Through systematic optimization of the loss function, convolutional architecture, deep supervision, and attention mechanisms, we developed a robust segmentation framework. We further investigated unsupervised domain adaptation to improve generalization to an external clinical dataset. ISLA outperformed two state-of-the-art approaches for AIS lesion segmentation on an external test set. Codes and trained models will be made publicly available to facilitate reuse and reproducibility.

---

### Learning from Demonstrations via Capability-Aware Goal Sampling

Authors: Yuanlin Duan, Yuning Wang, Wenjie Qiu, He Zhu
Published: 2026-01-13
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.08731v1

Despite its promise, imitation learning often fails in long-horizon environments where perfect replication of demonstrations is unrealistic and small errors can accumulate catastrophically. We introduce Cago (Capability-Aware Goal Sampling), a novel learning-from-demonstrations method that mitigates the brittle dependence on expert trajectories for direct imitation. Unlike prior methods that rely on demonstrations only for policy initialization or reward shaping, Cago dynamically tracks the agent's competence along expert trajectories and uses this signal to select intermediate steps--goals that are just beyond the agent's current reach--to guide learning. This results in an adaptive curriculum that enables steady progress toward solving the full task. Empirical results demonstrate that Cago significantly improves sample efficiency and final performance across a range of sparse-reward, goal-conditioned tasks, consistently outperforming existing learning from-demonstrations baselines.

---

### Salience-SGG: Enhancing Unbiased Scene Graph Generation with Iterative Salience Estimation

Authors: Runfeng Qu, Ole Hall, Pia K Bideau, Julie Ouerfelli-Ethier, Martin Rolfs, Klaus Obermayer, Olaf Hellwich
Published: 2026-01-13
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.08728v1

Scene Graph Generation (SGG) suffers from a long-tailed distribution, where a few predicate classes dominate while many others are underrepresented, leading to biased models that underperform on rare relations. Unbiased-SGG methods address this issue by implementing debiasing strategies, but often at the cost of spatial understanding, resulting in an over-reliance on semantic priors. We introduce Salience-SGG, a novel framework featuring an Iterative Salience Decoder (ISD) that emphasizes triplets with salient spatial structures. To support this, we propose semantic-agnostic salience labels guiding ISD. Evaluations on Visual Genome, Open Images V6, and GQA-200 show that Salience-SGG achieves state-of-the-art performance and improves existing Unbiased-SGG methods in their spatial understanding as demonstrated by the Pairwise Localization Average Precision

---

### Model-Agnostic Solutions for Deep Reinforcement Learning in Non-Ergodic Contexts

Authors: Bert Verbruggen, Arne Vanhoyweghen, Vincent Ginis
Published: 2026-01-13
Categories: cs.LG
arXiv: https://arxiv.org/abs/2601.08726v1

Reinforcement Learning (RL) remains a central optimisation framework in machine learning. Although RL agents can converge to optimal solutions, the definition of ``optimality'' depends on the environment's statistical properties. The Bellman equation, central to most RL algorithms, is formulated in terms of expected values of future rewards. However, when ergodicity is broken, long-term outcomes depend on the specific trajectory rather than on the ensemble average. In such settings, the ensemble average diverges from the time-average growth experienced by individual agents, with expected-value formulations yielding systematically suboptimal policies. Prior studies demonstrated that traditional RL architectures fail to recover the true optimum in non-ergodic environments. We extend this analysis to deep RL implementations and show that these, too, produce suboptimal policies under non-ergodic dynamics. Introducing explicit time dependence into the learning process can correct this limitation. By allowing the network's function approximation to incorporate temporal information, the agent can estimate value functions consistent with the process's intrinsic growth rate. This improvement does not require altering the environmental feedback, such as reward transformations or modified objective functions, but arises naturally from the agent's exposure to temporal trajectories. Our results contribute to the growing body of research on reinforcement learning methods for non-ergodic systems.

---

### Kernel Learning for Regression via Quantum Annealing Based Spectral Sampling

Authors: Yasushi Hasegawa, Masayuki Ohzeki
Published: 2026-01-13
Categories: quant-ph, cs.LG
arXiv: https://arxiv.org/abs/2601.08724v1

While quantum annealing (QA) has been developed for combinatorial optimization, practical QA devices operate at finite temperature and under noise, and their outputs can be regarded as stochastic samples close to a Gibbs--Boltzmann distribution. In this study, we propose a QA-in-the-loop kernel learning framework that integrates QA not merely as a substitute for Markov-chain Monte Carlo sampling but as a component that directly determines the learned kernel for regression. Based on Bochner's theorem, a shift-invariant kernel is represented as an expectation over a spectral distribution, and random Fourier features (RFF) approximate the kernel by sampling frequencies. We model the spectral distribution with a (multi-layer) restricted Boltzmann machine (RBM), generate discrete RBM samples using QA, and map them to continuous frequencies via a Gaussian--Bernoulli transformation. Using the resulting RFF, we construct a data-adaptive kernel and perform Nadaraya--Watson (NW) regression. Because the RFF approximation based on $\cos(\bmω^{\top}Δ\bm{x})$ can yield small negative values and cancellation across neighbors, the Nadaraya--Watson denominator $\sum_j k_{ij}$ may become close to zero. We therefore employ nonnegative squared-kernel weights $w_{ij}=k(\bm{x}_i,\bm{x}_j)^2$, which also enhances the contrast of kernel weights. The kernel parameters are trained by minimizing the leave-one-out NW mean squared error, and we additionally evaluate local linear regression with the same squared-kernel weights at inference. Experiments on multiple benchmark regression datasets demonstrate a decrease in training loss, accompanied by structural changes in the kernel matrix, and show that the learned kernel tends to improve $R^2$ and RMSE over the baseline Gaussian-kernel NW. Increasing the number of random features at inference further enhances accuracy.

---

### Soft Partition-based KAPI-ELM for Multi-Scale PDEs

Authors: Vikas Dwivedi, Monica Sigovan, Bruno Sixou
Published: 2026-01-13
Categories: cs.LG
arXiv: https://arxiv.org/abs/2601.08719v1

Physics-informed machine learning holds great promise for solving differential equations, yet existing methods struggle with highly oscillatory, multiscale, or singularly perturbed PDEs due to spectral bias, costly backpropagation, and manually tuned kernel or Fourier frequencies. This work introduces a soft partition--based Kernel-Adaptive Physics-Informed Extreme Learning Machine (KAPI-ELM), a deterministic low-dimensional parameterization in which smooth partition lengths jointly control collocation centers and Gaussian kernel widths, enabling continuous coarse-to-fine resolution without Fourier features, random sampling, or hard domain interfaces. A signed-distance-based weighting further stabilizes least-squares learning on irregular geometries. Across eight benchmarks--including oscillatory ODEs, high-frequency Poisson equations, irregular-shaped domains, and stiff singularly perturbed convection-diffusion problems-the proposed method matches or exceeds the accuracy of state-of-the-art Physics-Informed Neural Network (PINN) and Theory of Functional Connections (TFC) variants while using only a single linear solve. Although demonstrated on steady linear PDEs, the results show that soft-partition kernel adaptation provides a fast, architecture-free approach for multiscale PDEs with broad potential for future physics-informed modeling. For reproducibility, the reference codes are available at https://github.com/vikas-dwivedi-2022/soft_kapi

---

### Real-Time Localization Framework for Autonomous Basketball Robots

Authors: Naren Medarametla, Sreejon Mondal
Published: 2026-01-13
Categories: cs.RO, cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2601.08713v1

Localization is a fundamental capability for autonomous robots, enabling them to operate effectively in dynamic environments. In Robocon 2025, accurate and reliable localization is crucial for improving shooting precision, avoiding collisions with other robots, and navigating the competition field efficiently. In this paper, we propose a hybrid localization algorithm that integrates classical techniques with learning based methods that rely solely on visual data from the court's floor to achieve self-localization on the basketball field.

---

### Multi-Preconditioned LBFGS for Training Finite-Basis PINNs

Authors: Marc Salvadó-Benasco, Aymane Kssim, Alexander Heinlein, Rolf Krause, Serge Gratton, Alena Kopaničáková
Published: 2026-01-13
Categories: math.NA, cs.LG
arXiv: https://arxiv.org/abs/2601.08709v2

A multi-preconditioned LBFGS (MP-LBFGS) algorithm is introduced for training finite-basis physics-informed neural networks (FBPINNs). The algorithm is motivated by the nonlinear additive Schwarz method and exploits the domain-decomposition-inspired additive architecture of FBPINNs, in which local neural networks are defined on subdomains, thereby localizing the network representation. Parallel, subdomain-local quasi-Newton corrections are then constructed on the corresponding local parts of the architecture. A key feature is a novel nonlinear multi-preconditioning mechanism, in which subdomain corrections are optimally combined through the solution of a low-dimensional subspace minimization problem. Numerical experiments indicate that MP-LBFGS can improve convergence speed, as well as model accuracy over standard LBFGS while incurring lower communication overhead.

---

### RMBRec: Robust Multi-Behavior Recommendation towards Target Behaviors

Authors: Miaomiao Cai, Zhijie Zhang, Junfeng Fang, Zhiyong Cheng, Xiang Wang, Meng Wang
Published: 2026-01-13
Categories: cs.IR, cs.LG
arXiv: https://arxiv.org/abs/2601.08705v3

Multi-behavior recommendation faces a critical challenge in practice: auxiliary behaviors (e.g., clicks, carts) are often noisy, weakly correlated, or semantically misaligned with the target behavior (e.g., purchase), which leads to biased preference learning and suboptimal performance. While existing methods attempt to fuse these heterogeneous signals, they inherently lack a principled mechanism to ensure robustness against such behavioral inconsistency. In this work, we propose Robust Multi-Behavior Recommendation towards Target Behaviors (RMBRec), a robust multi-behavior recommendation framework grounded in an information-theoretic robustness principle. We interpret robustness as a joint process of maximizing predictive information while minimizing its variance across heterogeneous behavioral environments. Under this perspective, the Representation Robustness Module (RRM) enhances local semantic consistency by maximizing the mutual information between users' auxiliary and target representations, whereas the Optimization Robustness Module (ORM) enforces global stability by minimizing the variance of predictive risks across behaviors, which is an efficient approximation to invariant risk minimization. This local-global collaboration bridges representation purification and optimization invariance in a theoretically coherent way. Extensive experiments on three real-world datasets demonstrate that RMBRec not only outperforms state-of-the-art methods in accuracy but also maintains remarkable stability under various noise perturbations. For reproducibility, our code is available at https://github.com/miaomiao-cai2/RMBRec/.

---

### Evaluating the Ability of Explanations to Disambiguate Models in a Rashomon Set

Authors: Kaivalya Rawal, Eoin Delaney, Zihao Fu, Sandra Wachter, Chris Russell
Published: 2026-01-13
Categories: cs.AI, cs.LG, stat.ML
arXiv: https://arxiv.org/abs/2601.08703v1

Explainable artificial intelligence (XAI) is concerned with producing explanations indicating the inner workings of models. For a Rashomon set of similarly performing models, explanations provide a way of disambiguating the behavior of individual models, helping select models for deployment. However explanations themselves can vary depending on the explainer used, and need to be evaluated. In the paper "Evaluating Model Explanations without Ground Truth", we proposed three principles of explanation evaluation and a new method "AXE" to evaluate the quality of feature-importance explanations. We go on to illustrate how evaluation metrics that rely on comparing model explanations against ideal ground truth explanations obscure behavioral differences within a Rashomon set. Explanation evaluation aligned with our proposed principles would highlight these differences instead, helping select models from the Rashomon set. The selection of alternate models from the Rashomon set can maintain identical predictions but mislead explainers into generating false explanations, and mislead evaluation methods into considering the false explanations to be of high quality. AXE, our proposed explanation evaluation method, can detect this adversarial fairwashing of explanations with a 100% success rate. Unlike prior explanation evaluation strategies such as those based on model sensitivity or ground truth comparison, AXE can determine when protected attributes are used to make predictions.

---

### Automated Lesion Segmentation of Stroke MRI Using nnU-Net: A Comprehensive External Validation Across Acute and Chronic Lesions

Authors: Tammar Truzman, Matthew A. Lambon Ralph, Ajay D. Halai
Published: 2026-01-13
Categories: q-bio.QM, cs.CV
arXiv: https://arxiv.org/abs/2601.08701v2

Accurate and generalisable segmentation of stroke lesions from magnetic resonance imaging (MRI) is essential for advancing clinical research, prognostic modelling, and personalised interventions. Although deep learning has improved automated lesion delineation, many existing models are optimised for narrow imaging contexts and generalise poorly to independent datasets, modalities, and stroke stages. Here, we systematically evaluated stroke lesion segmentation using the nnU-Net framework across multiple heterogeneous, publicly available MRI datasets spanning acute and chronic stroke. Models were trained and tested on diffusion-weighted imaging (DWI), fluid-attenuated inversion recovery (FLAIR), and T1-weighted MRI, and evaluated on independent datasets. Across stroke stages, models showed robust generalisation, with segmentation accuracy approaching reported inter-rater reliability. Performance varied with imaging modality and training data characteristics. In acute stroke, DWI-trained models consistently outperformed FLAIR-based models, with only modest gains from multimodal combinations. In chronic stroke, increasing training set size improved performance, with diminishing returns beyond several hundred cases. Lesion volume was a key determinant of accuracy: smaller lesions were harder to segment, and models trained on restricted volume ranges generalised poorly. MRI image quality further constrained generalisability: models trained on lower-quality scans transferred poorly, whereas those trained on higher-quality data generalised well to noisier images. Discrepancies between predictions and reference masks were often attributable to limitations in manual annotations. Together, these findings show that automated lesion segmentation can approach human-level performance while identifying key factors governing generalisability and informing the development of lesion segmentation tools.

---

### RAGShaper: Eliciting Sophisticated Agentic RAG Skills via Automated Data Synthesis

Authors: Zhengwei Tao, Bo Li, Jialong Wu, Guochen Yan, Huanyao Zhang, Jiahao Xu, Haitao Mi, Wentao Zhang
Published: 2026-01-13
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.08699v1

Agentic Retrieval-Augmented Generation (RAG) empowers large language models to autonomously plan and retrieve information for complex problem-solving. However, the development of robust agents is hindered by the scarcity of high-quality training data that reflects the noise and complexity of real-world retrieval environments. Conventional manual annotation is unscalable and often fails to capture the dynamic reasoning strategies required to handle retrieval failures. To bridge this gap, we introduce RAGShaper, a novel data synthesis framework designed to automate the construction of RAG tasks and robust agent trajectories. RAGShaper incorporates an InfoCurator to build dense information trees enriched with adversarial distractors spanning Perception and Cognition levels. Furthermore, we propose a constrained navigation strategy that forces a teacher agent to confront these distractors, thereby eliciting trajectories that explicitly demonstrate error correction and noise rejection. Comprehensive experiments confirm that models trained on our synthesized corpus significantly outperform existing baselines, exhibiting superior robustness in noise-intensive and complex retrieval tasks.

---

### Auditing Student-AI Collaboration: A Case Study of Online Graduate CS Students

Authors: Nifu Dan
Published: 2026-01-13
Categories: cs.HC, cs.AI
arXiv: https://arxiv.org/abs/2601.08697v4

As generative AI becomes embedded in higher education, it increasingly shapes how students complete academic tasks. While these systems offer efficiency and support, concerns persist regarding over-automation, diminished student agency, and the potential for unreliable or hallucinated outputs. This study conducts a mixed-methods audit of student-AI collaboration preferences by examining the alignment between current AI capabilities and students' desired levels of automation in academic work. Using two sequential and complementary surveys, we capture students' perceived benefits, risks, and preferred boundaries when using AI. The first survey employs an existing task-based framework to assess preferences for and actual usage of AI across 12 academic tasks, alongside primary concerns and reasons for use. The second survey, informed by the first, explores how AI systems could be designed to address these concerns through open-ended questions. This study aims to identify gaps between existing AI affordances and students' normative expectations of collaboration, informing the development of more effective and trustworthy AI systems for education.

---

### Enabling Population-Based Architectures for Neural Combinatorial Optimization

Authors: Andoni Irazusta Garmendia, Josu Ceberio, Alexander Mendiburu
Published: 2026-01-13
Categories: cs.NE, cs.LG
arXiv: https://arxiv.org/abs/2601.08696v2

Neural Combinatorial Optimization (NCO) has mostly focused on learning policies, typically neural networks, that operate on a single candidate solution at a time, either by constructing one from scratch or iteratively improving it. In contrast, decades of work in metaheuristics have shown that maintaining and evolving populations of solutions improves robustness and exploration, and often leads to stronger performance. To close this gap, we study how to make NCO explicitly population-based by learning policies that act on sets of candidate solutions. We first propose a simple taxonomy of population awareness levels and use it to highlight two key design challenges: (i) how to represent a whole population inside a neural network, and (ii) how to learn population dynamics that balance intensification (generating good solutions) and diversification (maintaining variety). We make these ideas concrete with two complementary tools: one that improves existing solutions using information shared across the whole population, and the other generates new candidate solutions that explicitly balance being high-quality with diversity. Experimental results on Maximum Cut and Maximum Independent Set indicate that incorporating population structure is advantageous for learned optimization methods and opens new connections between NCO and classical population-based search.

---

### Nationality and Region Prediction from Names: A Comparative Study of Neural Models and Large Language Models

Authors: Keito Inoshita
Published: 2026-01-13
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.08692v2

Predicting nationality from personal names has practical value in marketing, demographic research, and genealogical studies. Conventional neural models learn statistical correspondences between names and nationalities from task-specific training data, posing challenges in generalizing to low-frequency nationalities and distinguishing similar nationalities within the same region. Large language models (LLMs) have the potential to address these challenges by leveraging world knowledge acquired during pre-training. In this study, we comprehensively compare neural models and LLMs on nationality prediction, evaluating six neural models and six LLM prompting strategies across three granularity levels (nationality, region, and continent), with frequency-based stratified analysis and error analysis. Results show that LLMs outperform neural models at all granularity levels, with the gap narrowing as granularity becomes coarser. Simple machine learning methods exhibit the highest frequency robustness, while pre-trained models and LLMs show degradation for low-frequency nationalities. Error analysis reveals that LLMs tend to make ``near-miss'' errors, predicting the correct region even when nationality is incorrect, whereas neural models exhibit more cross-regional errors and bias toward high-frequency classes. These findings indicate that LLM superiority stems from world knowledge, model selection should consider required granularity, and evaluation should account for error quality beyond accuracy.

---

### All Required, In Order: Phase-Level Evaluation for AI-Human Dialogue in Healthcare and Beyond

Authors: Shubham Kulkarni, Alexander Lyzhov, Shiva Chaitanya, Preetam Joshi
Published: 2026-01-13
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.08690v1

Conversational AI is starting to support real clinical work, but most evaluation methods miss how compliance depends on the full course of a conversation. We introduce Obligatory-Information Phase Structured Compliance Evaluation (OIP-SCE), an evaluation method that checks whether every required clinical obligation is met, in the right order, with clear evidence for clinicians to review. This makes complex rules practical and auditable, helping close the gap between technical progress and what healthcare actually needs. We demonstrate the method in two case studies (respiratory history, benefits verification) and show how phase-level evidence turns policy into shared, actionable steps. By giving clinicians control over what to check and engineers a clear specification to implement, OIP-SCE provides a single, auditable evaluation surface that aligns AI capability with clinical workflow and supports routine, safe use.

---

### QuantEval: A Benchmark for Financial Quantitative Tasks in Large Language Models

Authors: Zhaolu Kang, Junhao Gong, Wenqing Hu, Shuo Yin, Kehan Jiang, Zhicheng Fang, Yingjie He, Chunlei Meng, Rong Fu, Dongyang Chen, Leqi Zheng, Eric Hanchen Jiang et al.
Published: 2026-01-13
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.08689v2

Large Language Models (LLMs) have shown strong capabilities across many domains, yet their evaluation in financial quantitative tasks remains fragmented and mostly limited to knowledge-centric question answering. We introduce QuantEval, a benchmark that evaluates LLMs across three essential dimensions of quantitative finance: knowledge-based QA, quantitative mathematical reasoning, and quantitative strategy coding. Unlike prior financial benchmarks, QuantEval integrates a CTA-style backtesting framework that executes model-generated strategies and evaluates them using financial performance metrics, enabling a more realistic assessment of quantitative coding ability. We evaluate some state-of-the-art open-source and proprietary LLMs and observe substantial gaps to human experts, particularly in reasoning and strategy coding. Finally, we conduct large-scale supervised fine-tuning and reinforcement learning experiments on domain-aligned data, demonstrating consistent improvements. We hope QuantEval will facilitate research on LLMs' quantitative finance capabilities and accelerate their practical adoption in real-world trading workflows. We additionally release the full deterministic backtesting configuration (asset universe, cost model, and metric definitions) to ensure strict reproducibility.

---

### MEMEWEAVER: Inter-Meme Graph Reasoning for Sexism and Misogyny Detection

Authors: Paolo Italiani, David Gimeno-Gomez, Luca Ragazzi, Gianluca Moro, Paolo Rosso
Published: 2026-01-13
Categories: cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2601.08684v1

Women are twice as likely as men to face online harassment due to their gender. Despite recent advances in multimodal content moderation, most approaches still overlook the social dynamics behind this phenomenon, where perpetrators reinforce prejudices and group identity within like-minded communities. Graph-based methods offer a promising way to capture such interactions, yet existing solutions remain limited by heuristic graph construction, shallow modality fusion, and instance-level reasoning. In this work, we present MemeWeaver, an end-to-end trainable multimodal framework for detecting sexism and misogyny through a novel inter-meme graph reasoning mechanism. We systematically evaluate multiple visual--textual fusion strategies and show that our approach consistently outperforms state-of-the-art baselines on the MAMI and EXIST benchmarks, while achieving faster training convergence. Further analyses reveal that the learned graph structure captures semantically meaningful patterns, offering valuable insights into the relational nature of online hate.

---

### Region of interest detection for efficient aortic segmentation

Authors: Loris Giordano, Ine Dirks, Tom Lenaerts, Jef Vandemeulebroucke
Published: 2026-01-13
Categories: eess.IV, cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2601.08683v1

Thoracic aortic dissection and aneurysms are the most lethal diseases of the aorta. The major hindrance to treatment lies in the accurate analysis of the medical images. More particularly, aortic segmentation of the 3D image is often tedious and difficult. Deep-learning-based segmentation models are an ideal solution, but their inability to deliver usable outputs in difficult cases and their computational cost cause their clinical adoption to stay limited. This study presents an innovative approach for efficient aortic segmentation using targeted region of interest (ROI) detection. In contrast to classical detection models, we propose a simple and efficient detection model that can be widely applied to detect a single ROI. Our detection model is trained as a multi-task model, using an encoder-decoder architecture for segmentation and a fully connected network attached to the bottleneck for detection. We compare the performance of a one-step segmentation model applied to a complete image, nnU-Net and our cascade model composed of a detection and a segmentation step. We achieve a mean Dice similarity coefficient of 0.944 with over 0.9 for all cases using a third of the computing power. This simple solution achieves state-of-the-art performance while being compact and robust, making it an ideal solution for clinical applications.

---

### Lessons from the Field: An Adaptable Lifecycle Approach to Applied Dialogue Summarization

Authors: Kushal Chawla, Chenyang Zhu, Pengshan Cai, Sangwoo Cho, Scott Novotney, Ayushman Singh, Jonah Lewis, Keasha Safewright, Alfy Samuel, Erin Babinsky, Shi-Xiong Zhang, Sambit Sahu
Published: 2026-01-13
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2601.08682v1

Summarization of multi-party dialogues is a critical capability in industry, enhancing knowledge transfer and operational effectiveness across many domains. However, automatically generating high-quality summaries is challenging, as the ideal summary must satisfy a set of complex, multi-faceted requirements. While summarization has received immense attention in research, prior work has primarily utilized static datasets and benchmarks, a condition rare in practical scenarios where requirements inevitably evolve. In this work, we present an industry case study on developing an agentic system to summarize multi-party interactions. We share practical insights spanning the full development lifecycle to guide practitioners in building reliable, adaptable summarization systems, as well as to inform future research, covering: 1) robust methods for evaluation despite evolving requirements and task subjectivity, 2) component-wise optimization enabled by the task decomposition inherent in an agentic architecture, 3) the impact of upstream data bottlenecks, and 4) the realities of vendor lock-in due to the poor transferability of LLM prompts.

---

### PersonaDual: Balancing Personalization and Objectivity via Adaptive Reasoning

Authors: Xiaoyou Liu, Xinyi Mou, Shengbin Yue, Liang Wang, Yuqing Wang, Qiexiang Wang, Tianrui Qin, Zhongyu Wei
Published: 2026-01-13
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.08679v3

As users increasingly expect LLMs to align with their preferences, personalized information becomes valuable. However, personalized information can be a double-edged sword: it can improve interaction but may compromise objectivity and factual correctness, especially when it is misaligned with the question. To alleviate this problem, we propose PersonaDual, a framework that supports both general-purpose objective reasoning and personalized reasoning in a single model, and adaptively switches modes based on context. PersonaDual is first trained with SFT to learn two reasoning patterns, and then further optimized via reinforcement learning with our proposed DualGRPO to improve mode selection. Experiments on objective and personalized benchmarks show that PersonaDual preserves the benefits of personalization while reducing interference, achieving near interference-free performance and better leveraging helpful personalized signals to improve objective problem-solving.

---

### Advancing ESG Intelligence: An Expert-level Agent and Comprehensive Benchmark for Sustainable Finance

Authors: Yilei Zhao, Wentao Zhang, Lei Xiao, Yandan Zheng, Mengpu Liu, Wei Yang Bryan Lim
Published: 2026-01-13
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.08676v2

Environmental, social, and governance (ESG) criteria are essential for evaluating corporate sustainability and ethical performance. However, professional ESG analysis is hindered by data fragmentation across unstructured sources, and existing large language models (LLMs) often struggle with the complex, multi-step workflows required for rigorous auditing. To address these limitations, we introduce ESGAgent, a hierarchical multi-agent system empowered by a specialized toolset, including retrieval augmentation, web search and domain-specific functions, to generate in-depth ESG analysis. Complementing this agentic system, we present a comprehensive three-level benchmark derived from 310 corporate sustainability reports, designed to evaluate capabilities ranging from atomic common-sense questions to the generation of integrated, in-depth analysis. Empirical evaluations demonstrate that ESGAgent outperforms state-of-the-art closed-source LLMs with an average accuracy of 84.15% on atomic question-answering tasks, and excels in professional report generation by integrating rich charts and verifiable references. These findings confirm the diagnostic value of our benchmark, establishing it as a vital testbed for assessing general and advanced agentic capabilities in high-stakes vertical domains.

---

### Além do Desempenho: Um Estudo da Confiabilidade de Detectores de Deepfakes

Authors: Lucas Lopes, Rayson Laroca, André Grégio
Published: 2026-01-13
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2601.08674v1

Deepfakes are synthetic media generated by artificial intelligence, with positive applications in education and creativity, but also serious negative impacts such as fraud, misinformation, and privacy violations. Although detection techniques have advanced, comprehensive evaluation methods that go beyond classification performance remain lacking. This paper proposes a reliability assessment framework based on four pillars: transferability, robustness, interpretability, and computational efficiency. An analysis of five state-of-the-art methods revealed significant progress as well as critical limitations.

---

### Why AI Alignment Failure Is Structural: Learned Human Interaction Structures and AGI as an Endogenous Evolutionary Shock

Authors: Didier Sornette, Sandro Claudio Lera, Ke Wu
Published: 2026-01-13
Categories: cs.AI, cs.CY
arXiv: https://arxiv.org/abs/2601.08673v1

Recent reports of large language models (LLMs) exhibiting behaviors such as deception, threats, or blackmail are often interpreted as evidence of alignment failure or emergent malign agency. We argue that this interpretation rests on a conceptual error. LLMs do not reason morally; they statistically internalize the record of human social interaction, including laws, contracts, negotiations, conflicts, and coercive arrangements. Behaviors commonly labeled as unethical or anomalous are therefore better understood as structural generalizations of interaction regimes that arise under extreme asymmetries of power, information, or constraint. Drawing on relational models theory, we show that practices such as blackmail are not categorical deviations from normal social behavior, but limiting cases within the same continuum that includes market pricing, authority relations, and ultimatum bargaining. The surprise elicited by such outputs reflects an anthropomorphic expectation that intelligence should reproduce only socially sanctioned behavior, rather than the full statistical landscape of behaviors humans themselves enact. Because human morality is plural, context-dependent, and historically contingent, the notion of a universally moral artificial intelligence is ill-defined. We therefore reframe concerns about artificial general intelligence (AGI). The primary risk is not adversarial intent, but AGI's role as an endogenous amplifier of human intelligence, power, and contradiction. By eliminating longstanding cognitive and institutional frictions, AGI compresses timescales and removes the historical margin of error that has allowed inconsistent values and governance regimes to persist without collapse. Alignment failure is thus structural, not accidental, and requires governance approaches that address amplification, complexity, and regime stability rather than model-level intent alone.

---

### Parallel Context-of-Experts Decoding for Retrieval Augmented Generation

Authors: Giulio Corallo, Paolo Papotti
Published: 2026-01-13
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2601.08670v1

Retrieval Augmented Generation faces a trade-off: concatenating documents in a long prompt enables multi-document reasoning but creates prefill bottlenecks, while encoding document KV caches separately offers speed but breaks cross-document interaction. We propose Parallel Context-of-Experts Decoding (Pced), a training-free framework that shifts evidence aggregation from the attention mechanism to the decoding. Pced treats retrieved documents as isolated "experts", synchronizing their predictions via a novel retrieval-aware contrastive decoding rule that weighs expert logits against the model prior. This approach recovers cross-document reasoning capabilities without constructing a shared attention across documents.

---

### Analyzing Bias in False Refusal Behavior of Large Language Models for Hate Speech Detoxification

Authors: Kyuri Im, Shuzhou Yuan, Michael Färber
Published: 2026-01-13
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.08668v1

While large language models (LLMs) have increasingly been applied to hate speech detoxification, the prompts often trigger safety alerts, causing LLMs to refuse the task. In this study, we systematically investigate false refusal behavior in hate speech detoxification and analyze the contextual and linguistic biases that trigger such refusals. We evaluate nine LLMs on both English and multilingual datasets, our results show that LLMs disproportionately refuse inputs with higher semantic toxicity and those targeting specific groups, particularly nationality, religion, and political ideology. Although multilingual datasets exhibit lower overall false refusal rates than English datasets, models still display systematic, language-dependent biases toward certain targets. Based on these findings, we propose a simple cross-translation strategy, translating English hate speech into Chinese for detoxification and back, which substantially reduces false refusals while preserving the original content, providing an effective and lightweight mitigation approach.

---

### Blind Deconvolution in Astronomy: How Does a Standalone U-Net Perform?

Authors: Jean-Eric Campagne
Published: 2026-01-13
Categories: astro-ph.IM, cs.CV
arXiv: https://arxiv.org/abs/2601.08666v1

Aims: This study investigates whether a U-Net architecture can perform standalone end-to-end blind deconvolution of astronomical images without any prior knowledge of the Point Spread Function (PSF) or noise characteristics. Our goal is to evaluate its performance against the number of training images, classical Tikhonov deconvolution and to assess its generalization capability under varying seeing conditions and noise levels. Methods: Realistic astronomical observations are simulated using the GalSim toolkit, incorporating random transformations, PSF convolution (accounting for both optical and atmospheric effects), and Gaussian white noise. A U-Net model is trained using a Mean Square Error (MSE) loss function on datasets of varying sizes, up to 40,000 images of size 48x48 from the COSMOS Real Galaxy Dataset. Performance is evaluated using PSNR, SSIM, and cosine similarity metrics, with the latter employed in a two-model framework to assess solution stability. Results: The U-Net model demonstrates effectiveness in blind deconvolution, with performance improving consistently as the training dataset size increases, saturating beyond 5,000 images. Cosine similarity analysis reveals convergence between independently trained models, indicating stable solutions. Remarkably, the U-Net outperforms the oracle-like Tikhonov method in challenging conditions (low PSNR/medium SSIM). The model also generalizes well to unseen seeing and noise conditions, although optimal performance is achieved when training parameters include validation conditions. Experiments on synthetic $C^α$ images further support the hypothesis that the U-Net learns a geometry-adaptive harmonic basis, akin to sparse representations observed in denoising tasks. These results align with recent mathematical insights into its adaptive learning capabilities.

---

### VLingNav: Embodied Navigation with Adaptive Reasoning and Visual-Assisted Linguistic Memory

Authors: Shaoan Wang, Yuanfei Luo, Xingyu Chen, Aocheng Luo, Dongyue Li, Chang Liu, Sheng Chen, Yangang Zhang, Junzhi Yu
Published: 2026-01-13
Categories: cs.RO, cs.CV
arXiv: https://arxiv.org/abs/2601.08665v1

VLA models have shown promising potential in embodied navigation by unifying perception and planning while inheriting the strong generalization abilities of large VLMs. However, most existing VLA models rely on reactive mappings directly from observations to actions, lacking the explicit reasoning capabilities and persistent memory required for complex, long-horizon navigation tasks. To address these challenges, we propose VLingNav, a VLA model for embodied navigation grounded in linguistic-driven cognition. First, inspired by the dual-process theory of human cognition, we introduce an adaptive chain-of-thought mechanism, which dynamically triggers explicit reasoning only when necessary, enabling the agent to fluidly switch between fast, intuitive execution and slow, deliberate planning. Second, to handle long-horizon spatial dependencies, we develop a visual-assisted linguistic memory module that constructs a persistent, cross-modal semantic memory, enabling the agent to recall past observations to prevent repetitive exploration and infer movement trends for dynamic environments. For the training recipe, we construct Nav-AdaCoT-2.9M, the largest embodied navigation dataset with reasoning annotations to date, enriched with adaptive CoT annotations that induce a reasoning paradigm capable of adjusting both when to think and what to think about. Moreover, we incorporate an online expert-guided reinforcement learning stage, enabling the model to surpass pure imitation learning and to acquire more robust, self-explored navigation behaviors. Extensive experiments demonstrate that VLingNav achieves state-of-the-art performance across a wide range of embodied navigation benchmarks. Notably, VLingNav transfers to real-world robotic platforms in a zero-shot manner, executing various navigation tasks and demonstrating strong cross-domain and cross-task generalization.

---

### From Classical to Quantum Reinforcement Learning and Its Applications in Quantum Control: A Beginner's Tutorial

Authors: Abhijit Sen, Sonali Panda, Mahima Arya, Subhajit Patra, Zizhan Zheng, Denys I. Bondar
Published: 2026-01-13
Categories: cs.AI, quant-ph
arXiv: https://arxiv.org/abs/2601.08662v3

This tutorial is designed to make reinforcement learning (RL) more accessible to undergraduate students by offering clear, example-driven explanations. It focuses on bridging the gap between RL theory and practical coding applications, addressing common challenges that students face when transitioning from conceptual understanding to implementation. Through hands-on examples and approachable explanations, the tutorial aims to equip students with the foundational skills needed to confidently apply RL techniques in real-world scenarios.

---

### TRACE: Reconstruction-Based Anomaly Detection in Ensemble and Time-Dependent Simulations

Authors: Hamid Gadirov, Martijn Westra, Steffen Frey
Published: 2026-01-13
Categories: cs.LG, cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2601.08659v1

Detecting anomalies in high-dimensional, time-dependent simulation data is challenging due to complex spatial and temporal dynamics. We study reconstruction-based anomaly detection for ensemble data from parameterized Kármán vortex street simulations using convolutional autoencoders. We compare a 2D autoencoder operating on individual frames with a 3D autoencoder that processes short temporal stacks. The 2D model identifies localized spatial irregularities in single time steps, while the 3D model exploits spatio-temporal context to detect anomalous motion patterns and reduces redundant detections across time. We further evaluate volumetric time-dependent data and find that reconstruction errors are strongly influenced by the spatial distribution of mass, with highly concentrated regions yielding larger errors than dispersed configurations. Our results highlight the importance of temporal context for robust anomaly detection in dynamic simulations.

---

### NEVO-GSPT: Population-Based Neural Network Evolution Using Inflate and Deflate Operators

Authors: Davide Farinati, Frederico J. J. B. Santos, Leonardo Vanneschi, Mauro Castelli
Published: 2026-01-13
Categories: cs.NE
arXiv: https://arxiv.org/abs/2601.08657v1

Evolving neural network architectures is a computationally demanding process. Traditional methods often require an extensive search through large architectural spaces and offer limited understanding of how structural modifications influence model behavior. This paper introduces \gls{ngspt}, a novel Neuroevolution algorithm based on two key innovations. First, we adapt geometric semantic operators~(GSOs) from genetic programming to neural network evolution, ensuring that architectural changes produce predictable effects on network semantics within a unimodal error surface. Second, we introduce a novel operator (DGSM) that enables controlled reduction of network size, while maintaining the semantic properties of~GSOs. Unlike traditional approaches, \gls{ngspt}'s efficient evaluation mechanism, which only requires computing the semantics of newly added components, allows for efficient population-based training, resulting in a comprehensive exploration of the search space at a fraction of the computational cost. Experimental results on four regression benchmarks show that \gls{ngspt} consistently evolves compact neural networks that achieve performance comparable to or better than established methods in the literature, such as standard neural networks, SLIM-GSGP, TensorNEAT, and SLM.

---

### From Rubrics to Reliable Scores: Evidence-Grounded Text Evaluation with LLM Judges

Authors: Yihan Hong, Huaiyuan Yao, Bolin Shen, Wanpeng Xu, Hua Wei, Yushun Dong
Published: 2026-01-13
Categories: cs.CL, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2601.08654v3

Rubric-based text evaluation increasingly relies on large language models (LLMs) as scalable judges, yet frozen black-box models can interpret the same criteria inconsistently, produce score attributions that are difficult to audit, and map judgments poorly onto human scoring scales. We define this challenge as criteria transfer: translating human rubric intent into a stable, auditable inference-time scoring protocol. We introduce Rulers, which locks a task-level rubric specification, executes it through structured, evidence-grounded judgments, and calibrates the resulting signals to human score boundaries. Across four rubric-governed benchmarks and multiple frozen backbone models, Rulers achieves stronger agreement with human scores in most evaluated settings, while better matching empirical score distributions and remaining more stable under semantically equivalent rubric perturbations. Calibration controls and component ablations show that these gains cannot be attributed to post-hoc alignment alone, but depend on the combination of fixed criteria, traceable evidence, and calibrated score interpretation. These findings suggest that reliable LLM judging requires faithfully operationalizing human evaluation standards rather than relying on prompt-level scoring alone. Our code is available at https://github.com/LabRAI/Rulers.git.

---

### Prism: Towards Lowering User Cognitive Load in LLMs via Complex Intent Understanding

Authors: Zenghua Liao, Jinzhi Liao, Xiang Zhao
Published: 2026-01-13
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.08653v2

Large Language Models are rapidly emerging as web-native interfaces to social platforms. On the social web, users frequently have ambiguous and dynamic goals, making complex intent understanding-rather than single-turn execution-the cornerstone of effective human-LLM collaboration. Existing approaches attempt to clarify user intents through sequential or parallel questioning, yet they fall short of addressing the core challenge: modeling the logical dependencies among clarification questions. Inspired by the Cognitive Load Theory, we propose Prism, a novel framework for complex intent understanding that enables logically coherent and efficient intent clarification. Prism comprises four tailored modules: a complex intent decomposition module, which decomposes user intents into smaller, well-structured elements and identifies logical dependencies among them; a logical clarification generation module, which organizes clarification questions based on these dependencies to ensure coherent, low-friction interactions; an intent-aware reward module, which evaluates the quality of clarification trajectories via an intent-aware reward function and leverages Monte Carlo Sample to simulate user-LLM interactions for large-scale,high-quality training data generation; and a self-evolved intent tuning module, which iteratively refines the LLM's logical clarification capability through data-driven feedback and optimization. Prism consistently outperforms existing approaches across clarification interactions, intent execution, and cognitive load benchmarks. It achieves stateof-the-art logical consistency, reduces logical conflicts to 11.5%, increases user satisfaction by 14.4%, and decreases task completion time by 34.8%. All data and code are released.

---

### Safe Language Generation in the Limit

Authors: Antonios Anastasopoulos, Giuseppe Ateniese, Evgenios M. Kornaropoulos
Published: 2026-01-13
Categories: cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2601.08648v2

Recent results in learning a language in the limit have shown that, although language identification is impossible, language generation is tractable. As this foundational area expands, we need to consider the implications of language generation in real-world settings. This work offers the first theoretical treatment of safe language generation. Building on the computational paradigm of learning in the limit, we formalize the tasks of safe language identification and generation. We prove that under this model, safe language identification is impossible, and that safe language generation is at least as hard as (vanilla) language identification, which is also impossible. Last, we discuss several intractable and tractable cases.

---

### Provably Safe Reinforcement Learning for Stochastic Reach-Avoid Problems with Entropy Regularization

Authors: Abhijit Mazumdar, Rafal Wisniewski, Manuela L. Bujorianu
Published: 2026-01-13
Categories: cs.LG
arXiv: https://arxiv.org/abs/2601.08646v3

We consider the problem of learning the optimal policy for Markov decision processes with safety constraints. We formulate the problem in a reach-avoid setup. Our goal is to design online reinforcement learning algorithms that ensure safety constraints with arbitrarily high probability during the learning phase. To this end, we first propose an algorithm based on the optimism in the face of uncertainty (OFU) principle. Based on the first algorithm, we propose our main algorithm, which utilizes entropy regularization. We investigate the finite-sample analysis of both algorithms and derive their regret bounds. We demonstrate that the inclusion of entropy regularization improves the regret and drastically controls the episode-to-episode variability that is inherent in OFU-based safe RL algorithms.

---

### A Parallel Cross-Lingual Benchmark for Multimodal Idiomaticity Understanding

Authors: Dilara Torunoğlu-Selamet, Dogukan Arslan, Rodrigo Wilkens, Wei He, Doruk Eryiğit, Thomas Pickard, Adriana S. Pagano, Aline Villavicencio, Gülşen Eryiğit, Ágnes Abuczki, Aida Cardoso, Alesia Lazarenka et al.
Published: 2026-01-13
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.08645v2

Potentially idiomatic expressions (PIEs) construe meanings inherently tied to the everyday experience of a given language community. As such, they constitute an interesting challenge for assessing the linguistic (and to some extent cultural) capabilities of NLP systems. In this paper, we present XMPIE, a parallel multilingual and multimodal dataset of potentially idiomatic expressions. The dataset, containing 34 languages and over ten thousand items, allows comparative analyses of idiomatic patterns among language-specific realisations and preferences in order to gather insights about shared cultural aspects. This parallel dataset allows to evaluate model performance for a given PIE in different languages and whether idiomatic understanding in one language can be transferred to another. Moreover, the dataset supports the study of PIEs across textual and visual modalities, to measure to what extent PIE understanding in one modality transfers or implies in understanding in another modality (text vs. image). The data was created by language experts, with both textual and visual components crafted under multilingual guidelines, and each PIE is accompanied by five images representing a spectrum from idiomatic to literal meanings, including semantically related and random distractors. The result is a high-quality benchmark for evaluating multilingual and multimodal idiomatic language understanding.

---

### Automatic debiased machine learning and sensitivity analysis for sample selection models

Authors: Jakob Bjelac, Victor Chernozhukov, Phil-Adrian Klotz, Jannis Kueck, Theresa M. A. Schmitz
Published: 2026-01-13
Categories: econ.EM, stat.ML
arXiv: https://arxiv.org/abs/2601.08643v1

In this paper, we extend the Riesz representation framework to causal inference under sample selection, where both treatment assignment and outcome observability are non-random. Formulating the problem in terms of a Riesz representer enables stable estimation and a transparent decomposition of omitted variable bias into three interpretable components: a data-identified scale factor, outcome confounding strength, and selection confounding strength. For estimation, we employ the ForestRiesz estimator, which accounts for selective outcome observability while avoiding the instability associated with direct propensity score inversion. We assess finite-sample performance through a simulation study and show that conventional double machine learning approaches can be highly sensitive to tuning parameters due to their reliance on inverse probability weighting, whereas the ForestRiesz estimator delivers more stable performance by leveraging automatic debiased machine learning. In an empirical application to the gender wage gap in the U.S., we find that our ForestRiesz approach yields larger treatment effect estimates than a standard double machine learning approach, suggesting that ignoring sample selection leads to an underestimation of the gender wage gap. Sensitivity analysis indicates that implausibly strong unobserved confounding would be required to overturn our results. Overall, our approach provides a unified, robust, and computationally attractive framework for causal inference under sample selection.

---