Tootfinder

Differentially Private aggregate hints in mev-share
Jonathan Passerat-Palmbach, Sarisht Wadhwa
https://arxiv.org/abs/2508.14284 https://arxiv.org/pdf/2508.…

Differentially Private aggregate hints in mev-share
Flashbots recently released mev-share to empower users with control over the amount of information they share with searchers for extracting Maximal Extractable Value (MEV). Searchers require more information to maintain on-chain exchange efficiency and profitability, while users aim to prevent frontrunning by withholding information. After analyzing two searching strategies in mev-share to reason about searching techniques, this paper introduces Differentially-Private (DP) aggregate hints as a …

FLAIR: Feedback Learning for Adaptive Information Retrieval
William Zhang, Yiwen Zhu, Yunlei Lu, Mathieu Demarne, Wenjing Wang, Kai Deng, Nutan Sahoo, Katherine Lin, Miso Cilimdzic, Subru Krishnan
https://arxiv.org/abs/2508.13390

FLAIR: Feedback Learning for Adaptive Information Retrieval
Recent advances in Large Language Models (LLMs) have driven the adoption of copilots in complex technical scenarios, underscoring the growing need for specialized information retrieval solutions. In this paper, we introduce FLAIR, a lightweight, feedback learning framework that adapts copilot systems' retrieval strategies by integrating domain-specific expert feedback. FLAIR operates in two stages: an offline phase obtains indicators from (1) user feedback and (2) questions synthesized from doc…

Tokenization Strategies for Low-Resource Agglutinative Languages in Word2Vec: Case Study on Turkish and Finnish
Jinfan Frank Hu
https://arxiv.org/abs/2509.14238 https://

Tokenization Strategies for Low-Resource Agglutinative Languages in Word2Vec: Case Study on Turkish and Finnish
Tokenization plays a critical role in processing agglutinative languages, where a single word can encode multiple morphemes carrying syntactic and semantic information. This study evaluates the impact of various tokenization strategies - word-level, character-level, n-gram, and Byte Pair Encoding (BPE) - on the quality of static word embeddings generated by Word2Vec for Turkish and Finnish. Using a 10,000-article Wikipedia corpus, we trained models under low-resource conditions and evaluated th…

Breaking the SFT Plateau: Multimodal Structured Reinforcement Learning for Chart-to-Code Generation
Lei Chen, Xuanle Zhao, Zhixiong Zeng, Jing Huang, Liming Zheng, Yufeng Zhong, Lin Ma
https://arxiv.org/abs/2508.13587

Breaking the SFT Plateau: Multimodal Structured Reinforcement Learning for Chart-to-Code Generation
While reinforcement learning (RL) has proven highly effective for general reasoning in vision-language models, its application to tasks requiring in-depth understanding of information-rich images and generation of structured outputs remains underexplored. Chart-to-code generation exemplifies this challenge, demanding complex reasoning over visual charts to generate structured code. Supervised fine-tuning (SFT) alone is often insufficient, highlighting the need for effective RL strategies that a…

AFABench: A Generic Framework for Benchmarking Active Feature Acquisition
Valter Sch\"utz, Han Wu, Reza Rezvan, Linus Aronsson, Morteza Haghir Chehreghani
https://arxiv.org/abs/2508.14734

AFABench: A Generic Framework for Benchmarking Active Feature Acquisition
In many real-world scenarios, acquiring all features of a data instance can be expensive or impractical due to monetary cost, latency, or privacy concerns. Active Feature Acquisition (AFA) addresses this challenge by dynamically selecting a subset of informative features for each data instance, trading predictive performance against acquisition cost. While numerous methods have been proposed for AFA, ranging from greedy information-theoretic strategies to non-myopic reinforcement learning appro…

Sensing the Shape of Data: Non-Visual Exploration of Statistical Concepts in Histograms with Blind and Low-Vision Learners
Sanchita S. Kamath, Omar Khan, Aziz N Zeidieh, JooYoung Seo
https://arxiv.org/abs/2509.14452

Sensing the Shape of Data: Non-Visual Exploration of Statistical Concepts in Histograms with Blind and Low-Vision Learners
Statistical concepts often rely heavily on visual cues for comprehension, presenting challenges for individuals who face difficulties using visual information, such as the blind and low-vision (BLV) community. While prior work has explored making data visualizations accessible, limited research examines how BLV individuals conceptualize and learn the underlying statistical concepts these visualizations represent. To better understand BLV individuals' learning strategies for potentially unfamili…

LAURON VI: A Six-Legged Robot for Dynamic Walking
Christian Eichmann (FZI Research Center for Information Technology), Sabine Bellmann (FZI Research Center for Information Technology), Nicolas H\"ugel (FZI Research Center for Information Technology), Louis-Elias Enslin (FZI Research Center for Information Technology), Carsten Plasberg (FZI Research Center for Information Technology), Georg Heppner (FZI Research Center for Information Technology), Arne Roennau (FZI Research Center …

LAURON VI: A Six-Legged Robot for Dynamic Walking
Legged locomotion enables robotic systems to traverse extremely challenging terrains. In many real-world scenarios, the terrain is not that difficult and these mixed terrain types introduce the need for flexible use of different walking strategies to achieve mission goals in a fast, reliable, and energy-efficient way. Six-legged robots have a high degree of flexibility and inherent stability that aids them in traversing even some of the most difficult terrains, such as collapsed buildings. Howe…

On Finite-Blocklength Noisy Classical-Quantum Channel Coding With Amplitude Damping Errors
Tam\'as Havas, Hsuan-Yin Lin, Eirik Rosnes, Ching-Yi Lai
https://arxiv.org/abs/2509.14852

On Finite-Blocklength Noisy Classical-Quantum Channel Coding With Amplitude Damping Errors
We investigate practical finite-blocklength classical-quantum channel coding over the quantum amplitude damping channel (ADC), aiming to transmit classical information reliably through quantum outputs. Our findings indicate that for any finite blocklength, a naive (uncoded) approach fails to offer any advantage over the ADC. Instead, sophisticated encoding strategies that leverage both classical error-correcting codes and quantum input states are crucial for realizing quantum performance gains …

Strategies educators can use to counter misinformation related to the quantum information revolution
Jaya Shivangani Kashyap, Chandralekha Singh
https://arxiv.org/abs/2509.12489

Strategies educators can use to counter misinformation related to the quantum information revolution
Remarkable advances in quantum information science and technology (QIST) have taken place in recent years. However, they have also been accompanied by widespread misinformation. This paper provides suggestions for how educators can help students at all levels and especially early learners including those at the pre-college and college levels learn key QIST concepts so that they are less likely to be misinformed, e.g., by online unvetted resources. We discuss findings from interviews with five c…

How Quantum Agents Can Change Which Strategies Are More Complex
Spiros Kechrimparis, Nix Barnett, Mile Gu, Hyukjoon Kwon
https://arxiv.org/abs/2508.08092 https://

How Quantum Agents Can Change Which Strategies Are More Complex
Whether winning blackjack or navigating busy streets, achieving desired outcomes requires agents to execute adaptive strategies, strategies where actions depend contextually on past events. In complexity science, this motivates memory as an operational quantifier of complexity: given two strategies, the more complex one demands the agent to track more about the past. Here, we show that conclusions about complexity fundamentally depend on whether agents can process and store quantum information.…

"Central to Fletcher’s thesis is the concept of 'story thinking.' This cognitive framework aligns closely with how humans have evolved to process information and communicate experiences. He posits that storytelling is [...] a fundamental cognitive tool that shapes our actions and decisions. The ability to conceptualize new narratives allows humans to forge paths toward success, whether by escaping predators or devising innovative strategies in modern contexts".

Toxicity in State Sponsored Information Operations
Ashfaq Ali Shafin, Khandaker Mamun Ahmed
https://arxiv.org/abs/2507.10936 https://…

Toxicity in State Sponsored Information Operations
State-sponsored information operations (IOs) increasingly influence global discourse on social media platforms, yet their emotional and rhetorical strategies remain inadequately characterized in scientific literature. This study presents the first comprehensive analysis of toxic language deployment within such campaigns, examining 56 million posts from over 42 thousand accounts linked to 18 distinct geopolitical entities on X/Twitter. Using Google's Perspective API, we systematically detect and…

TableDART: Dynamic Adaptive Multi-Modal Routing for Table Understanding
Xiaobo Xing, Wei Yuan, Tong Chen, Quoc Viet Hung Nguyen, Xiangliang Zhang, Hongzhi Yin
https://arxiv.org/abs/2509.14671

TableDART: Dynamic Adaptive Multi-Modal Routing for Table Understanding
Modeling semantic and structural information from tabular data remains a core challenge for effective table understanding. Existing Table-as-Text approaches flatten tables for large language models (LLMs), but lose crucial structural cues, while Table-as-Image methods preserve structure yet struggle with fine-grained semantics. Recent Table-as-Multimodality strategies attempt to combine textual and visual views, but they (1) statically process both modalities for every query-table pair within a…

INFNet: A Task-aware Information Flow Network for Large-Scale Recommendation Systems
Kaiyuan Li, Dongdong Mao, Yongxiang Tang, Yanhua Cheng, Yanxiang Zeng, Chao Wang, Xialong Liu, Peng Jiang
https://arxiv.org/abs/2508.11565

INFNet: A Task-aware Information Flow Network for Large-Scale Recommendation Systems
Feature interaction has long been a cornerstone of ranking models in large-scale recommender systems due to its proven effectiveness in capturing complex dependencies among features. However, existing feature interaction strategies face two critical challenges in industrial applications: (1) The vast number of categorical and sequential features makes exhaustive interaction computationally prohibitive, often resulting in optimization difficulties. (2) Real-world recommender systems typically in…

Conditional Information Bottleneck for Multimodal Fusion: Overcoming Shortcut Learning in Sarcasm Detection
Yihua Wang, Qi Jia, Cong Xu, Feiyu Chen, Yuhan Liu, Haotian Zhang, Liang Jin, Lu Liu, Zhichun Wang
https://arxiv.org/abs/2508.10644

Conditional Information Bottleneck for Multimodal Fusion: Overcoming Shortcut Learning in Sarcasm Detection
Multimodal sarcasm detection is a complex task that requires distinguishing subtle complementary signals across modalities while filtering out irrelevant information. Many advanced methods rely on learning shortcuts from datasets rather than extracting intended sarcasm-related features. However, our experiments show that shortcut learning impairs the model's generalization in real-world scenarios. Furthermore, we reveal the weaknesses of current modality fusion strategies for multimodal sarcasm…

JFRFFNet: A Data-Model Co-Driven Graph Signal Denoising Model with Partial Prior Information
Ziqi Yan, Zhichao Zhang
https://arxiv.org/abs/2509.09147 https://

JFRFFNet: A Data-Model Co-Driven Graph Signal Denoising Model with Partial Prior Information
Wiener filtering in the joint time-vertex fractional Fourier transform (JFRFT) domain has shown high effectiveness in denoising time-varying graph signals. Traditional filtering models use grid search to determine the transform-order pair and compute filter coefficients, while learnable ones employ gradient-descent strategies to optimize them; both require complete prior information of graph signals. To overcome this shortcoming, this letter proposes a data-model co-driven denoising approach, t…

Hyperspectral Mamba for Hyperspectral Object Tracking
Long Gao, Yunhe Zhang, Yan Jiang, Weiying Xie, Yunsong Li
https://arxiv.org/abs/2509.08265 https://ar…

Hyperspectral Mamba for Hyperspectral Object Tracking
Hyperspectral object tracking holds great promise due to the rich spectral information and fine-grained material distinctions in hyperspectral images, which are beneficial in challenging scenarios. While existing hyperspectral trackers have made progress by either transforming hyperspectral data into false-color images or incorporating modality fusion strategies, they often fail to capture the intrinsic spectral information, temporal dependencies, and cross-depth interactions. To address these …

In-between Transatlantic (Monetary) Disturbances
Santiago Camara, Jeanne Aublin
https://arxiv.org/abs/2509.13578 https://arxiv.org/pdf/2509.13578

In-between Transatlantic (Monetary) Disturbances
This paper studies the spillovers of European Central Bank (ECB) interest rate shocks into the Canadian economy and compares them with those of the U.S. Federal Reserve (Fed). We combine a VAR model and local projection regressions with identification strategies that explicitly purge information effects around policy announcements. We find that an ECB rate hike leads to a depreciation of the Canadian dollar and a sharp contraction in economic activity. The main transmission channel is internati…

Analyzing Information-Seeking Behaviors in a Hakka AI Chatbot: A Cognitive-Pragmatic Study
Chu-Hsuan Lee, Chen-Chi Chang, Hung-Shin Lee, Yun-Hsiang Hsu, Ching-Yuan Chen
https://arxiv.org/abs/2509.11591

Analyzing Information-Seeking Behaviors in a Hakka AI Chatbot: A Cognitive-Pragmatic Study
With many endangered languages at risk of disappearing, efforts to preserve them now rely more than ever on using technology alongside culturally informed teaching strategies. This study examines user behaviors in TALKA, a generative AI-powered chatbot designed for Hakka language engagement, by employing a dual-layered analytical framework grounded in Bloom's Taxonomy of cognitive processes and dialogue act categorization. We analyzed 7,077 user utterances, each carefully annotated according to…

DFed-SST: Building Semantic- and Structure-aware Topologies for Decentralized Federated Graph Learning
Lianshuai Guo, Zhongzheng Yuan, Xunkai Li, Yinlin Zhu, Meixia Qu, Wenyu Wang
https://arxiv.org/abs/2508.11530

DFed-SST: Building Semantic- and Structure-aware Topologies for Decentralized Federated Graph Learning
Decentralized Federated Learning (DFL) has emerged as a robust distributed paradigm that circumvents the single-point-of-failure and communication bottleneck risks of centralized architectures. However, a significant challenge arises as existing DFL optimization strategies, primarily designed for tasks such as computer vision, fail to address the unique topological information inherent in the local subgraph. Notably, while Federated Graph Learning (FGL) is tailored for graph data, it is predomi…

Role-Augmented Intent-Driven Generative Search Engine Optimization
Xiaolu Chen, Haojie Wu, Jie Bao, Zhen Chen, Yong Liao, Hu Huang
https://arxiv.org/abs/2508.11158 https://

Role-Augmented Intent-Driven Generative Search Engine Optimization
Generative Search Engines (GSEs), powered by Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG), are reshaping information retrieval. While commercial systems (e.g., BingChat, Perplexity.ai) demonstrate impressive semantic synthesis capabilities, their black-box nature fundamentally undermines established Search Engine Optimization (SEO) practices. Content creators face a critical challenge: their optimization strategies, effective in traditional search engines, are misaligne…

Last-Iterate Convergence in Adaptive Regret Minimization for Approximate Extensive-Form Perfect Equilibrium
Hang Ren, Xiaozhen Sun, Tianzi Ma, Jiajia Zhang, Xuan Wang
https://arxiv.org/abs/2508.07699

Last-Iterate Convergence in Adaptive Regret Minimization for Approximate Extensive-Form Perfect Equilibrium
The Nash Equilibrium (NE) assumes rational play in imperfect-information Extensive-Form Games (EFGs) but fails to ensure optimal strategies for off-equilibrium branches of the game tree, potentially leading to suboptimal outcomes in practical settings. To address this, the Extensive-Form Perfect Equilibrium (EFPE), a refinement of NE, introduces controlled perturbations to model potential player errors. However, existing EFPE-finding algorithms, which typically rely on average strategy converge…

Digital Iran Reloaded: Gamer Mitigation Tactics of IRI Information Controls
Melinda Cohoon
https://arxiv.org/abs/2509.09063 https://arxiv.org/pdf/2509.0906…

Digital Iran Reloaded: Gamer Mitigation Tactics of IRI Information Controls
Internet censorship in the Islamic Republic of Iran restricts access to global platforms and services, forcing users to rely on circumvention technologies such as VPNs, proxies, and tunneling tools. This report presents findings from a mixed-methods study of 660 Iranian internet users, with a focus on gamers as a digitally literate and socially networked community. Survey data are combined with network measurements of latency and VPN performance to identify both technical and social strategies …

Searching for Privacy Risks in LLM Agents via Simulation
Yanzhe Zhang, Diyi Yang
https://arxiv.org/abs/2508.10880 https://arxiv.org/pdf/2508.10880

Searching for Privacy Risks in LLM Agents via Simulation
The widespread deployment of LLM-based agents is likely to introduce a critical privacy threat: malicious agents that proactively engage others in multi-turn interactions to extract sensitive information. These dynamic dialogues enable adaptive attack strategies that can cause severe privacy violations, yet their evolving nature makes it difficult to anticipate and discover sophisticated vulnerabilities manually. To tackle this problem, we present a search-based framework that alternates betwee…

Kahan's Automatic Step-Size Control for Unconstrained Optimization
Yifeng Meng, Chungen Shen, Linuo Xue, Lei-Hong Zhang
https://arxiv.org/abs/2508.06002 https://

Kahan's Automatic Step-Size Control for Unconstrained Optimization
The Barzilai and Borwein (BB) gradient method is one of the most widely-used line-search gradient methods. It computes the step-size for the current iterate by using the information carried in the previous iteration. Recently, William Kahan [Kahan, Automatic Step-Size Control for Minimization Iterations, Technical report, University of California, Berkeley CA, USA, 2019] proposed new Gradient Descent (KGD) step-size strategies which iterate the step-size itself by effectively utilizing the info…

Recovering link-weight structure in complex networks with weight-aware random walks
Adilson Vital Jr., Filipi N. Silva, Diego R. Amancio
https://arxiv.org/abs/2508.07489 https:/…

Recovering link-weight structure in complex networks with weight-aware random walks
Using edge weights is essential for modeling real-world systems where links possess relevant information, and preserving this information in low-dimensional representations is relevant for classification and prediction tasks. This paper systematically investigates how different random walk strategies - traditional unweighted, strength-based, and fully weight-aware - keeps edge weight information when generating node embeddings. Using network models, real-world graphs, and networks subjected to …

Differential Attention for Multimodal Crisis Event Analysis
Nusrat Munia, Junfeng Zhu, Olfa Nasraoui, Abdullah-Al-Zubaer Imran
https://arxiv.org/abs/2507.05165

Differential Attention for Multimodal Crisis Event Analysis
Social networks can be a valuable source of information during crisis events. In particular, users can post a stream of multimodal data that can be critical for real-time humanitarian response. However, effectively extracting meaningful information from this large and noisy data stream and effectively integrating heterogeneous data remains a formidable challenge. In this work, we explore vision language models (VLMs) and advanced fusion strategies to enhance the classification of crisis data in…

Training Strategies for Modality Dropout Resilient Multi-Modal Target Speaker Extraction
Srikanth Korse, Mohamed Elminshawi, Emanuel A. P. Habets, Srikanth Raj Chetupalli
https://arxiv.org/abs/2507.06566

Training Strategies for Modality Dropout Resilient Multi-Modal Target Speaker Extraction
The primary goal of multi-modal TSE (MTSE) is to extract a target speaker from a speech mixture using complementary information from different modalities, such as audio enrolment and visual feeds corresponding to the target speaker. MTSE systems are expected to perform well even when one of the modalities is unavailable. In practice, the systems often suffer from modality dominance, where one of the modalities outweighs the others, thereby limiting robustness. Our study investigates training st…

Contrastive Learning with Enhanced Abstract Representations using Grouped Loss of Abstract Semantic Supervision
Omri Suissa, Muhiim Ali, Shengmai Chen, Yinuo Cai, Shekhar Pradhan
https://arxiv.org/abs/2509.12771

Contrastive Learning with Enhanced Abstract Representations using Grouped Loss of Abstract Semantic Supervision
Humans can recognize an image as an instance of a general concept, beyond simply identifying its objects and their relationships. In this paper, we investigate 1. The extent to which VLMs have this concept abstraction capacity, and 2. Strategies for encoding the sort of higher-concept information in images that would enable the resulting VLM model (CLEAR GLASS model) to have this capability to a greater degree. To this end, we introduce a grouped image-caption dataset (MAGIC), which consists of…

Balancing Information Preservation and Disentanglement in Self-Supervised Music Representation Learning
Julia Wilkins, Sivan Ding, Magdalena Fuentes, Juan Pablo Bello
https://arxiv.org/abs/2507.22995

Balancing Information Preservation and Disentanglement in Self-Supervised Music Representation Learning
Recent advances in self-supervised learning (SSL) methods offer a range of strategies for capturing useful representations from music audio without the need for labeled data. While some techniques focus on preserving comprehensive details through reconstruction, others favor semantic structure via contrastive objectives. Few works examine the interaction between these paradigms in a unified SSL framework. In this work, we propose a multi-view SSL framework for disentangling music audio represen…

Maker playing against an invisible Breaker
Dennis Clemens, Fabian Hamann, Mirjana Mikala\v{c}ki, Yannick Mogge, Milo\v{s} Stojakovi\'c
https://arxiv.org/abs/2507.22519 https…

Maker playing against an invisible Breaker
We initiate the study of the phantom version of Maker-Breaker positional games. In a phantom game, the moves of one of the players are hidden from the other player, who still has the complete information. We look at the biased $(a:b)$ Maker-PhantomBreaker games where the board is the edge set of the complete graph on $n$ vertices, $K_n$, and Maker has no information about PhantomBreaker's choices of edges. We give randomized strategies for both players in four classical games: connectivity game…

Quantum MIMO Diversity over Discrete-Variable Crosstalk Channels
Shehbaz Tariq, Junaid ur Rehman, Symeon Chatzinotas
https://arxiv.org/abs/2508.07344 https://

Quantum MIMO Diversity over Discrete-Variable Crosstalk Channels
Quantum communication plays a pivotal role in enabling distributed quantum computing and sensing. Diversity strategies can be used to increase the communication reliability (in the sense of output fidelity with respect to the input quantum state) when multiple communication channels are available. The current paper proposes a diversity strategy for quantum discrete variable (DV) multiple-input-multiple-output (MIMO) channels, utilizing approximate cloning to distribute information across multip…

Memory-aware feedback enhances power in active information engines
Sehoon Bahng, Jae Sung Lee, Cheol-Min Ghim
https://arxiv.org/abs/2508.16866 https://arxi…

Memory-aware feedback enhances power in active information engines
We study an information engine operating in an active bath, where a Brownian particle confined in a harmonic trap undergoes feedback-driven displacement cycles. Unlike thermal environments, active baths exhibit temporally correlated fluctuations, introducing memory effects that challenge conventional feedback strategies. Extending the framework of stochastic thermodynamics to account for such memory, we analyze a feedback protocol that periodically shifts the potential minimum based on noisy me…

MM-DREX: Multimodal-Driven Dynamic Routing of LLM Experts for Financial Trading
Yang Chen, Yueheng Jiang, Zhaozhao Ma, Yuchen Cao Jacky Keung, Kun Kuang, Leilei Gan, Yiquan Wu, Fei Wu
https://arxiv.org/abs/2509.05080

MM-DREX: Multimodal-Driven Dynamic Routing of LLM Experts for Financial Trading
The inherent non-stationarity of financial markets and the complexity of multi-modal information pose significant challenges to existing quantitative trading models. Traditional methods relying on fixed structures and unimodal data struggle to adapt to market regime shifts, while large language model (LLM)-driven solutions - despite their multi-modal comprehension - suffer from static strategies and homogeneous expert designs, lacking dynamic adjustment and fine-grained decision mechanisms. To …

Learning-Enhanced Variational Regularization for Electrical Impedance Tomography via \Calderon's Method
Kai Li, Kwancheol Shin, Zhi Zhou
https://arxiv.org/abs/2507.06114

Learning-Enhanced Variational Regularization for Electrical Impedance Tomography via \Calderon's Method
This paper aims to numerically solve the two-dimensional electrical impedance tomography (EIT) with Cauchy data. This inverse problem is highly challenging due to its severe ill-posed nature and strong nonlinearity, which necessitates appropriate regularization strategies. Choosing a regularization approach that effectively incorporates the \textit{a priori} information of the conductivity distribution (or its contrast) is therefore essential. In this work, we propose a deep learning-based meth…

Uncharted Waters: Selling a New Product Robustly
Kun Zhang
https://arxiv.org/abs/2508.04134 https://arxiv.org/pdf/2508.04134…

Uncharted Waters: Selling a New Product Robustly
When introducing a novel product, a seller sets a price and decides how much information to provide to a buyer, who may incur a search cost to discover an outside option. The buyer knows the outside option distribution; the seller knows only its mean and bounds. Seeking "robustness," the seller evaluates strategies based on guaranteed profits, balancing search deterrence against surplus extraction. Providing information can deter search and boost demand but requires offering the buyer a higher …

Do Students Write Better Post-AI Support? Effects of Generative AI Literacy and Chatbot Interaction Strategies on Multimodal Academic Writing
Yueqiao Jin, Kaixun Yang, Roberto Martinez-Maldonado, Dragan Ga\v{s}evi\'c, Lixiang Yan
https://arxiv.org/abs/2507.04398

Do Students Write Better Post-AI Support? Effects of Generative AI Literacy and Chatbot Interaction Strategies on Multimodal Academic Writing
Academic writing increasingly involves multimodal tasks requiring students to integrate visual information and textual arguments. While generative AI (GenAI) tools, like ChatGPT, offer new pathways for supporting academic writing, little is known about how students' GenAI literacy influences their independent multimodal writing skills or how chatbot interaction strategies (passive reactive vs. proactive scaffolding) impact learning. This study examined 79 higher education students' multimodal a…

Noise or Nuance: An Investigation Into Useful Information and Filtering For LLM Driven AKBC
Alex Clay, Ernesto Jim\'enez-Ruiz, Pranava Madhyastha
https://arxiv.org/abs/2509.08903

Noise or Nuance: An Investigation Into Useful Information and Filtering For LLM Driven AKBC
RAG and fine-tuning are prevalent strategies for improving the quality of LLM outputs. However, in constrained situations, such as that of the 2025 LM-KBC challenge, such techniques are restricted. In this work we investigate three facets of the triple completion task: generation, quality assurance, and LLM response parsing. Our work finds that in this constrained setting: additional information improves generation quality, LLMs can be effective at filtering poor quality triples, and the tradeo…

Exciton transport driven by spin excitations in an antiferromagnet
Florian Dirnberger, Sophia Terres, Zakhar A. Iakovlev, Kseniia Mosina, Zdenek Sofer, Akashdeep Kamra, Mikhail M. Glazov, Alexey Chernikov
https://arxiv.org/abs/2507.07071

Exciton transport driven by spin excitations in an antiferromagnet
A new class of optical quasiparticles called magnetic excitons recently emerged in magnetic van der Waals materials. Akin to the highly effective strategies developed for electrons, the strong interactions of these excitons with the spin degree of freedom may provide innovative solutions for long-standing challenges in optics, such as steering the flow of energy and information. Here, we demonstrate transport of excitons by spin excitations in the van der Waals antiferromagnetic semiconductor C…

A Study of Large Language Models for Patient Information Extraction: Model Architecture, Fine-Tuning Strategy, and Multi-task Instruction Tuning
Cheng Peng, Xinyu Dong, Mengxian Lyu, Daniel Paredes, Yaoyun Zhang, Yonghui Wu
https://arxiv.org/abs/2509.04753

A Study of Large Language Models for Patient Information Extraction: Model Architecture, Fine-Tuning Strategy, and Multi-task Instruction Tuning
Natural language processing (NLP) is a key technology to extract important patient information from clinical narratives to support healthcare applications. The rapid development of large language models (LLMs) has revolutionized many NLP tasks in the clinical domain, yet their optimal use in patient information extraction tasks requires further exploration. This study examines LLMs' effectiveness in patient information extraction, focusing on LLM architectures, fine-tuning strategies, and multi…

Balancing utility and cost in dynamic treatment regimes
Kai Chen, Zhang Yuqian
https://arxiv.org/abs/2507.17360 https://arxiv.org/pdf/2507.17360

Balancing utility and cost in dynamic treatment regimes
Dynamic treatment regimes (DTRs) are personalized, adaptive strategies designed to guide the sequential allocation of treatments based on individual characteristics over time. Before each treatment assignment, covariate information is collected to refine treatment decisions and enhance their effectiveness. The more information we gather, the more precise our decisions can be. However, this also leads to higher costs during the data collection phase. In this work, we propose a balanced Q-learnin…

Quantum metrology through spectral measurements in quantum optics
Alejandro Vivas-Via\~na, Carlos S\'anchez Mu\~noz
https://arxiv.org/abs/2509.04300 https://

Quantum metrology through spectral measurements in quantum optics
Continuously monitored quantum systems are emerging as promising platforms for quantum metrology, where a central challenge is to identify measurement strategies that optimally extract information about unknown parameters encoded in the complex quantum state of emitted radiation. Different measurement strategies effectively access distinct temporal modes of the emitted field, and the resulting choice of mode can strongly impact the information available for parameter estimation. While a ubiquit…

Generative AI for Critical Infrastructure in Smart Grids: A Unified Framework for Synthetic Data Generation and Anomaly Detection
Aydin Zaboli, Junho Hong
https://arxiv.org/abs/2508.08593

Generative AI for Critical Infrastructure in Smart Grids: A Unified Framework for Synthetic Data Generation and Anomaly Detection
In digital substations, security events pose significant challenges to the sustained operation of power systems. To mitigate these challenges, the implementation of robust defense strategies is critically important. A thorough process of anomaly identification and detection in information and communication technology (ICT) frameworks is crucial to ensure secure and reliable communication and coordination between interconnected devices within digital substations. Hence, this paper addresses the …

Beating the Best Constant Rebalancing Portfolio in Long-Term Investment: A Generalization of the Kelly Criterion and Universal Learning Algorithm for Markets with Serial Dependence
Duy Khanh Lam
https://arxiv.org/abs/2507.05994

Beating the Best Constant Rebalancing Portfolio in Long-Term Investment: A Generalization of the Kelly Criterion and Universal Learning Algorithm for Markets with Serial Dependence
In the online portfolio optimization framework, existing learning algorithms generate strategies that yield significantly poorer cumulative wealth compared to the best constant rebalancing portfolio in hindsight, despite being consistent in asymptotic growth rate. While this unappealing performance can be improved by incorporating more side information, it raises difficulties in feature selection and high-dimensional settings. Instead, the inherent serial dependence of assets' returns, such as …

Robust Competitive Ratio for Deterministic Monopoly Pricing
Tim S. G. van Eck, Pieter Kleer, Johan S. H. van Leeuwaarden
https://arxiv.org/abs/2509.06619 https://

Robust Competitive Ratio for Deterministic Monopoly Pricing
We study deterministic monopoly pricing under partial knowledge of the market, where the seller has access only to summary statistics of the valuation distribution, such as the mean, dispersion, and maximum value. Using tools from distributionally robust optimization and max-min analysis, we evaluate pricing strategies based on their competitive ratio (CR). We characterize the worst-case market scenario consistent with the available information and provide a complete solution for minimizing the…

Optimal interventions in opinion dynamics on large-scale, time-varying, random networks
Leonardo Cianfanelli, Giacomo Como, Fabio Fagnani, Asuman Ozdaglar, Francesca Parise
https://arxiv.org/abs/2509.02623

Optimal interventions in opinion dynamics on large-scale, time-varying, random networks
We consider two optimization problems in which a planner aims to influence the average transient opinion in the Friedkin-Johnsen dynamics on a network by intervening on the agents' innate opinions. Solving these problems requires full network knowledge, which is often not available because of the cost involved in collecting this information or due to privacy considerations. For this reason, we focus on intervention strategies that are based on statistical instead of exact knowledge of the netwo…

Hypercomplex Prompt-aware Multimodal Recommendation
Zheyu Chen, Jinfeng Xu, Hewei Wang, Shuo Yang, Zitong Wan, Haibo Hu
https://arxiv.org/abs/2508.10753 https://

Hypercomplex Prompt-aware Multimodal Recommendation
Modern recommender systems face critical challenges in handling information overload while addressing the inherent limitations of multimodal representation learning. Existing methods suffer from three fundamental limitations: (1) restricted ability to represent rich multimodal features through a single representation, (2) existing linear modality fusion strategies ignore the deep nonlinear correlations between modalities, and (3) static optimization methods failing to dynamically mitigate the o…

Distillation versus Contrastive Learning: How to Train Your Rerankers
Zhichao Xu, Zhiqi Huang, Shengyao Zhuang, Ashim Gupta, Vivek Srikumar
https://arxiv.org/abs/2507.08336

Distillation versus Contrastive Learning: How to Train Your Rerankers
Training text rerankers is crucial for information retrieval. Two primary strategies are widely used: contrastive learning (optimizing directly on ground-truth labels) and knowledge distillation (transferring knowledge from a larger reranker). While both have been studied in the literature, a clear comparison of their effectiveness for training cross-encoder rerankers under practical conditions is needed. This paper empirically compares these strategies by training rerankers of different size…

Information Entropy-Based Scheduling for Communication-Efficient Decentralized Learning
Jaiprakash Nagar, Zheng Chen, Marios Kountouris, Photios A. Stavrou
https://arxiv.org/abs/2507.17426

Information Entropy-Based Scheduling for Communication-Efficient Decentralized Learning
This paper addresses decentralized stochastic gradient descent (D-SGD) over resource-constrained networks by introducing node-based and link-based scheduling strategies to enhance communication efficiency. In each iteration of the D-SGD algorithm, only a few disjoint subsets of nodes or links are randomly activated, subject to a given communication cost constraint. We propose a novel importance metric based on information entropy to determine node and link scheduling probabilities. We validate …

Beyond One Shot, Beyond One Perspective: Cross-View and Long-Horizon Distillation for Better LiDAR Representations
Xiang Xu, Lingdong Kong, Song Wang, Chuanwei Zhou, Qingshan Liu
https://arxiv.org/abs/2507.05260

Beyond One Shot, Beyond One Perspective: Cross-View and Long-Horizon Distillation for Better LiDAR Representations
LiDAR representation learning aims to extract rich structural and semantic information from large-scale, readily available datasets, reducing reliance on costly human annotations. However, existing LiDAR representation strategies often overlook the inherent spatiotemporal cues in LiDAR sequences, limiting their effectiveness. In this work, we propose LiMA, a novel long-term image-to-LiDAR Memory Aggregation framework that explicitly captures longer range temporal correlations to enhance LiDAR r…

Privacy-Preserving Distributed Control for a Networked Battery Energy Storage System
Mihitha Maithripala, Zongli Lin
https://arxiv.org/abs/2508.19345 https://

Privacy-Preserving Distributed Control for a Networked Battery Energy Storage System
The increasing deployment of distributed Battery Energy Storage Systems (BESSs) in modern power grids necessitates effective coordination strategies to ensure state-of-charge (SoC) balancing and accurate power delivery. While distributed control frameworks offer scalability and resilience, they also raise significant privacy concerns due to the need for inter-agent information exchange. This paper presents a novel privacy-preserving distributed control algorithm for SoC balancing in a networked…

Biaxialformer: Leveraging Channel Independence and Inter-Channel Correlations in EEG Signal Decoding for Predicting Neurological Outcomes
Naimahmed Nesaragi, Hemin Ali Qadir, Per Steiner Halvorsen, Ilangko Balasingham
https://arxiv.org/abs/2507.02879

Biaxialformer: Leveraging Channel Independence and Inter-Channel Correlations in EEG Signal Decoding for Predicting Neurological Outcomes
Accurate decoding of EEG signals requires comprehensive modeling of both temporal dynamics within individual channels and spatial dependencies across channels. While Transformer-based models utilizing channel-independence (CI) strategies have demonstrated strong performance in various time series tasks, they often overlook the inter-channel correlations that are critical in multivariate EEG signals. This omission can lead to information degradation and reduced prediction accuracy, particularly …

Pre-training Time Series Models with Stock Data Customization
Mengyu Wang, Tiejun Ma, Shay B. Cohen
https://arxiv.org/abs/2506.16746 https://

Pre-training Time Series Models with Stock Data Customization
Stock selection, which aims to predict stock prices and identify the most profitable ones, is a crucial task in finance. While existing methods primarily focus on developing model structures and building graphs for improved selection, pre-training strategies remain underexplored in this domain. Current stock series pre-training follows methods from other areas without adapting to the unique characteristics of financial data, particularly overlooking stock-specific contextual information and the…

Phishing Webpage Detection: Unveiling the Threat Landscape and Investigating Detection Techniques
Aditya Kulkarni, Vivek Balachandran, Tamal Das
https://arxiv.org/abs/2509.08424

Phishing Webpage Detection: Unveiling the Threat Landscape and Investigating Detection Techniques
In the realm of cybersecurity, phishing stands as a prevalent cyber attack, where attackers employ various tactics to deceive users into gathering their sensitive information, potentially leading to identity theft or financial gain. Researchers have been actively working on advancing phishing webpage detection approaches to detect new phishing URLs, bolstering user protection. Nonetheless, the ever-evolving strategies employed by attackers, aimed at circumventing existing detection approaches a…

Reliable Weak-to-Strong Monitoring of LLM Agents
Neil Kale, Chen Bo Calvin Zhang, Kevin Zhu, Ankit Aich, Paula Rodriguez, Scale Red Team, Christina Q. Knight, Zifan Wang
https://arxiv.org/abs/2508.19461

Reliable Weak-to-Strong Monitoring of LLM Agents
We stress test monitoring systems for detecting covert misbehavior in autonomous LLM agents (e.g., secretly sharing private information). To this end, we systematize a monitor red teaming (MRT) workflow that incorporates: (1) varying levels of agent and monitor situational awareness; (2) distinct adversarial strategies to evade the monitor, such as prompt injection; and (3) two datasets and environments -- SHADE-Arena for tool-calling agents and our new CUA-SHADE-Arena, which extends TheAgentCo…

Thinking Inside the Mask: In-Place Prompting in Diffusion LLMs
Xiangqi Jin, Yuxuan Wang, Yifeng Gao, Zichen Wen, Biqing Qi, Dongrui Liu, Linfeng Zhang
https://arxiv.org/abs/2508.10736

Thinking Inside the Mask: In-Place Prompting in Diffusion LLMs
Despite large language models (LLMs) have achieved remarkable success, their prefix-only prompting paradigm and sequential generation process offer limited flexibility for bidirectional information. Diffusion large language models (dLLMs) present new opportunities through their bidirectional attention mechanisms and iterative refinement processes, enabling more flexible in-place prompting strategies. We introduce ICE (In-Place Chain-of-Thought Prompting with Early Exit), a novel framework that …

Exploring Non-Local Spatial-Angular Correlations with a Hybrid Mamba-Transformer Framework for Light Field Super-Resolution
Haosong Liu, Xiancheng Zhu, Huanqiang Zeng, Jianqing Zhu, Jiuwen Cao, Junhui Hou
https://arxiv.org/abs/2509.04824

Exploring Non-Local Spatial-Angular Correlations with a Hybrid Mamba-Transformer Framework for Light Field Super-Resolution
Recently, Mamba-based methods, with its advantage in long-range information modeling and linear complexity, have shown great potential in optimizing both computational cost and performance of light field image super-resolution (LFSR). However, current multi-directional scanning strategies lead to inefficient and redundant feature extraction when applied to complex LF data. To overcome this challenge, we propose a Subspace Simple Scanning (Sub-SS) strategy, based on which we design the Subspace …

SaD: A Scenario-Aware Discriminator for Speech Enhancement
Xihao Yuan, Siqi Liu, Yan Chen, Hang Zhou, Chang Liu, Hanting Chen, Jie Hu
https://arxiv.org/abs/2509.00405 https://…

SaD: A Scenario-Aware Discriminator for Speech Enhancement
Generative adversarial network-based models have shown remarkable performance in the field of speech enhancement. However, the current optimization strategies for these models predominantly focus on refining the architecture of the generator or enhancing the quality evaluation metrics of the discriminator. This approach often overlooks the rich contextual information inherent in diverse scenarios. In this paper, we propose a scenario-aware discriminator that captures scene-specific features and…

Fast and Accurate SVD-Type Updating in Streaming Data
Johannes J. Brust, Michael A. Saunders
https://arxiv.org/abs/2509.02840 https://arxiv.org/pdf/2509.02…

Fast and Accurate SVD-Type Updating in Streaming Data
For a datastream, the change over a short interval is often of low rank. For high throughput information arranged in matrix format, recomputing an optimal SVD approximation after each step is typically prohibitive. Instead, incremental and truncated updating strategies are used, which may not scale for large truncation ranks. Therefore, we propose a set of efficient new algorithms that update a bidiagonal factorization, and which are similarly accurate as the SVD methods. In particular, we deve…

Acoustic Non-Stationarity Objective Assessment with Hard Label Criteria for Supervised Learning Models
Guilherme Zucatelli, Ricardo Barioni, Gabriela Dantas
https://arxiv.org/abs/2508.06405

Acoustic Non-Stationarity Objective Assessment with Hard Label Criteria for Supervised Learning Models
Objective non-stationarity measures are resource intensive and impose critical limitations for real-time processing solutions. In this paper, a novel Hard Label Criteria (HLC) algorithm is proposed to generate a global non-stationarity label for acoustic signals, enabling supervised learning strategies to be trained as stationarity estimators. The HLC is first evaluated on state-of-the-art general-purpose acoustic models, demonstrating that these models encode stationarity information. Furtherm…

Discrimination by LLMs: Cross-lingual Bias Assessment and Mitigation in Decision-Making and Summarisation
Willem Huijzer, Jieying Chen
https://arxiv.org/abs/2509.09735 https://

Discrimination by LLMs: Cross-lingual Bias Assessment and Mitigation in Decision-Making and Summarisation
The rapid integration of Large Language Models (LLMs) into various domains raises concerns about societal inequalities and information bias. This study examines biases in LLMs related to background, gender, and age, with a focus on their impact on decision-making and summarization tasks. Additionally, the research examines the cross-lingual propagation of these biases and evaluates the effectiveness of prompt-instructed mitigation strategies. Using an adapted version of the dataset by Tamkin et…

Unnoticeable Community Deception via Multi-objective Optimization
Junyuan Fang, Huimin Liu, Yueqi Peng, Jiajing Wu, Zibin Zheng, Chi K. Tse
https://arxiv.org/abs/2509.01438 http…

Unnoticeable Community Deception via Multi-objective Optimization
Community detection in graphs is crucial for understanding the organization of nodes into densely connected clusters. While numerous strategies have been developed to identify these clusters, the success of community detection can lead to privacy and information security concerns, as individuals may not want their personal information exposed. To address this, community deception methods have been proposed to reduce the effectiveness of detection algorithms. Nevertheless, several limitations, s…

LitForager: Exploring Multimodal Literature Foraging Strategies in Immersive Sensemaking
Haoyang Yang, Elliott H. Faa, Weijian Liu, Shunan Guo, Duen Horng Chau, Yalong Yang
https://arxiv.org/abs/2508.15043

LitForager: Exploring Multimodal Literature Foraging Strategies in Immersive Sensemaking
Exploring and comprehending relevant academic literature is a vital yet challenging task for researchers, especially given the rapid expansion in research publications. This task fundamentally involves sensemaking - interpreting complex, scattered information sources to build understanding. While emerging immersive analytics tools have shown cognitive benefits like enhanced spatial memory and reduced mental load, they predominantly focus on information synthesis (e.g., organizing known document…

Quantum Resource Correction
Mark Byrd, Daniel Dilley, Alvin Gonzales, Masaya Takahashi, Zain Saleem, Lian-Ao Wu
https://arxiv.org/abs/2506.19776 https://…

Quantum Resource Correction
Resource theories play a crucial role in characterizing states and properties essential for quantum information processing. A significant challenge is protecting resources from errors. We explore strategies for correcting quantum resources. We show that resource preserving operations in resource theory define a gauge freedom on code spaces, which allows for recovery strategies that can correct the resource while changing non-essential properties. This allows decoding to be simplified. The resul…

A Comprehensive Incremental and Ensemble Learning Approach for Forecasting Individual Electric Vehicle Charging Parameters
Parnian Alikhani, Nico Brinkel, Wouter Schram, Ioannis Lampropoulos, Wilfried van Sark
https://arxiv.org/abs/2508.17772

A Comprehensive Incremental and Ensemble Learning Approach for Forecasting Individual Electric Vehicle Charging Parameters
Electric vehicles (EVs) have the potential to reduce grid stress through smart charging strategies while simultaneously meeting user demand. This requires accurate forecasts of key charging parameters, such as energy demand and connection time. Although previous studies have made progress in this area, they have overlooked the importance of dynamic training to capture recent patterns and have excluded EV sessions with limited information, missing potential opportunities to use these data. To ad…

Evaluating Language Models For Threat Detection in IoT Security Logs
Jorge J. Tejero-Fern\'andez, Alfonso S\'anchez-Maci\'an
https://arxiv.org/abs/2507.02390

Evaluating Language Models For Threat Detection in IoT Security Logs
Log analysis is a relevant research field in cybersecurity as they can provide a source of information for the detection of threats to networks and systems. This paper presents a pipeline to use fine-tuned Large Language Models (LLMs) for anomaly detection and mitigation recommendation using IoT security logs. Utilizing classical machine learning classifiers as a baseline, three open-source LLMs are compared for binary and multiclass anomaly detection, with three strategies: zero-shot, few-shot…

Octo's Heartland: Supporting Children with Congenital Heart Disease through Digital Health Education
Irene Zeng, Neda Barbazi, Ji Youn Shin, Gurumurthy Hiremath, Carlye Anne Lauff
https://arxiv.org/abs/2509.02537

Octo's Heartland: Supporting Children with Congenital Heart Disease through Digital Health Education
Children with congenital heart disease (CHD) often face challenges that require them to understand complex medical information from an early age in order to support lifelong care and improve health outcomes. However, prior research has rarely included young children in designing and evaluating digital tools to support health education using developmentally appropriate strategies. This study is part of a multi-phase research involving participatory design (PD), user testing, and iterative develo…

Explainable speech emotion recognition through attentive pooling: insights from attention-based temporal localization
Tahitoa Leygue (DIASI), Astrid Sabourin (DIASI), Christian Bolzmacher (DIASI), Sylvain Bouchigny (DIASI), Margarita Anastassova (DIASI), Quoc-Cuong Pham (DIASI)
https://arxiv.org/abs/2506.15754

Explainable speech emotion recognition through attentive pooling: insights from attention-based temporal localization
State-of-the-art transformer models for Speech Emotion Recognition (SER) rely on temporal feature aggregation, yet advanced pooling methods remain underexplored. We systematically benchmark pooling strategies, including Multi-Query Multi-Head Attentive Statistics Pooling, which achieves a 3.5 percentage point macro F1 gain over average pooling. Attention analysis shows 15 percent of frames capture 80 percent of emotion cues, revealing a localized pattern of emotional information. Analysis of hi…

X-Troll: eXplainable Detection of State-Sponsored Information Operations Agents
Lin Tian, Xiuzhen Zhang, Maria Myung-Hee Kim, Jennifer Biggs, Marian-Andrei Rizoiu
https://arxiv.org/abs/2508.16021

X-Troll: eXplainable Detection of State-Sponsored Information Operations Agents
State-sponsored trolls, malicious actors who deploy sophisticated linguistic manipulation in coordinated information campaigns, posing threats to online discourse integrity. While Large Language Models (LLMs) achieve strong performance on general natural language processing (NLP) tasks, they struggle with subtle propaganda detection and operate as ``black boxes'', providing no interpretable insights into manipulation strategies. This paper introduces X-Troll, a novel framework that bridges this…

Empowering Iterative Graph Alignment Using Heat Diffusion
Boyan Wang, Weijie Feng, Jinyang Huang, Dan Guo, Zhi Liu
https://arxiv.org/abs/2506.17640 https:/…

Empowering Iterative Graph Alignment Using Heat Diffusion
Unsupervised plain graph alignment (UPGA) aims to align corresponding nodes across two graphs without any auxiliary information. Existing UPGA methods rely on structural consistency while neglecting the inherent structural differences in real-world graphs, leading to biased node representations. Moreover, their one-shot alignment strategies lack mechanisms to correct erroneous matches arising from inaccurate anchor seeds. To address these issues, this paper proposes IterAlign, a novel parameter…

Amplifying Machine Learning Attacks Through Strategic Compositions
Yugeng Liu, Zheng Li, Hai Huang, Michael Backes, Yang Zhang
https://arxiv.org/abs/2506.18870

Amplifying Machine Learning Attacks Through Strategic Compositions
Machine learning (ML) models are proving to be vulnerable to a variety of attacks that allow the adversary to learn sensitive information, cause mispredictions, and more. While these attacks have been extensively studied, current research predominantly focuses on analyzing each attack type individually. In practice, however, adversaries may employ multiple attack strategies simultaneously rather than relying on a single approach. This prompts a crucial yet underexplored question: When the adver…

Dynamic Simulation Framework for Disinformation Dissemination and Correction With Social Bots
Boyu Qiao, Kun Li, Wei Zhou, Songlin Hu
https://arxiv.org/abs/2507.16848 https://…

Dynamic Simulation Framework for Disinformation Dissemination and Correction With Social Bots
In the human-bot symbiotic information ecosystem, social bots play key roles in spreading and correcting disinformation. Understanding their influence is essential for risk control and better governance. However, current studies often rely on simplistic user and network modeling, overlook the dynamic behavior of bots, and lack quantitative evaluation of correction strategies. To fill these gaps, we propose MADD, a Multi Agent based framework for Disinformation Dissemination. MADD constructs a m…

Entropy-Driven Pre-Tokenization for Byte-Pair Encoding
Yifan Hu, Frank Liang, Dachuan Zhao, Jonathan Geuter, Varshini Reddy, Craig W. Schmidt, Chris Tanner
https://arxiv.org/abs/2506.15889

Entropy-Driven Pre-Tokenization for Byte-Pair Encoding
Byte-Pair Encoding (BPE) has become a widely adopted subword tokenization method in modern language models due to its simplicity and strong empirical performance across downstream tasks. However, applying BPE to unsegmented languages such as Chinese presents significant challenges, as its frequency-driven merge operation is agnostic to linguistic boundaries. To address this, we propose two entropy-informed pre-tokenization strategies that guide BPE segmentation using unsupervised information-th…