Tootfinder

Git Context Controller: Manage the Context of LLM-based Agents like Git
Junde Wu
https://arxiv.org/abs/2508.00031 https://arxiv.org/pdf/2508.00031

Git Context Controller: Manage the Context of LLM-based Agents like Git
Large language model (LLM) based agents have shown impressive capabilities by interleaving internal reasoning with external tool use. However, as these agents are deployed in long-horizon workflows, such as coding for a big, long-term project, context management becomes a critical bottleneck. We introduce Git-Context-Controller (GCC), a structured context management framework inspired by software version control systems. GCC elevates context as versioned memory hierarchy like Git. It structures…

Context as Memory: Scene-Consistent Interactive Long Video Generation with Memory Retrieval
Jiwen Yu, Jianhong Bai, Yiran Qin, Quande Liu, Xintao Wang, Pengfei Wan, Di Zhang, Xihui Liu
https://arxiv.org/abs/2506.03141

Context as Memory: Scene-Consistent Interactive Long Video Generation with Memory Retrieval
Recent advances in interactive video generation have shown promising results, yet existing approaches struggle with scene-consistent memory capabilities in long video generation due to limited use of historical context. In this work, we propose Context-as-Memory, which utilizes historical context as memory for video generation. It includes two simple yet effective designs: (1) storing context in frame format without additional post-processing; (2) conditioning by concatenating context and frame…

Variational Uncertainty Decomposition for In-Context Learning
I. Shavindra Jayasekera, Jacob Si, Wenlong Chen, Filippo Valdettaro, A. Aldo Faisal, Yingzhen Li
https://arxiv.org/abs/2509.02327

Variational Uncertainty Decomposition for In-Context Learning
As large language models (LLMs) gain popularity in conducting prediction tasks in-context, understanding the sources of uncertainty in in-context learning becomes essential to ensuring reliability. The recent hypothesis of in-context learning performing predictive Bayesian inference opens the avenue for Bayesian uncertainty estimation, particularly for decomposing uncertainty into epistemic uncertainty due to lack of in-context data and aleatoric uncertainty inherent in the in-context predictio…

A substitution lemma for multiple context-free languages
Andrew Duncan, Murray Elder, Lisa Frenkel, Mengfan Lyu
https://arxiv.org/abs/2509.02117 https://ar…

A substitution lemma for multiple context-free languages
We present a new criterion for proving that a language is not multiple context-free, which we call a Substitution Lemma. We apply it to show a sample selection of languages are not multiple context-free, including the word problem of $F_2\times F_2$. Our result is in contrast to Kanazawa et al. [2014, Theory Comput. Syst.] who proved that it was not possible to generalise the standard pumping lemma for context-free languages to multiple context-free languages, and Kanazawa [2019, Inform. and …

OMD at the Greek Theater in Los Angeles, performing 'Anthropocene' June 21, 2025.
https://open.spotify.com/track/1zTghGUFb7Ajrn5ghO9ZQl?si=XNxEhurBRXaWIu3kJmY42Q&context=spotify:album:2yW1JUK…

Anthropocene
Orchestral Manoeuvres In The Dark · Bauhaus Staircase (The Complete Edition) · Song · 2023

Disentangled Multi-Context Meta-Learning: Unlocking robust and Generalized Task Learning
Seonsoo Kim, Jun-Gill Kang, Taehong Kim, Seongil Hong
https://arxiv.org/abs/2509.01297 h…

Disentangled Multi-Context Meta-Learning: Unlocking robust and Generalized Task Learning
In meta-learning and its downstream tasks, many methods rely on implicit adaptation to task variations, where multiple factors are mixed together in a single entangled representation. This makes it difficult to interpret which factors drive performance and can hinder generalization. In this work, we introduce a disentangled multi-context meta-learning framework that explicitly assigns each task factor to a distinct context vector. By decoupling these variations, our approach improves robustness…

Out-of-Context Abduction: LLMs Make Inferences About Procedural Data Leveraging Declarative Facts in Earlier Training Data
Sohaib Imran, Rob Lamb, Peter M. Atkinson
https://arxiv.org/abs/2508.00741

Out-of-Context Abduction: LLMs Make Inferences About Procedural Data Leveraging Declarative Facts in Earlier Training Data
Large language models (LLMs) are trained on large corpora, yet it is unclear whether they can reason about the information present within their training data. We design experiments to study out-of-context abduction in LLMs, the ability to infer the most plausible explanations for observations using relevant facts present in training data. We train treatment LLMs on names and behavior descriptions of fictitious chatbots, but not on examples of dialogue with the chatbots. We find that OpenAI's GP…

Beyond the Protocol: Unveiling Attack Vectors in the Model Context Protocol Ecosystem
Hao Song, Yiming Shen, Wenxuan Luo, Leixin Guo, Ting Chen, Jiashui Wang, Beibei Li, Xiaosong Zhang, Jiachi Chen
https://arxiv.org/abs/2506.02040

Beyond the Protocol: Unveiling Attack Vectors in the Model Context Protocol Ecosystem
The Model Context Protocol (MCP) is an emerging standard designed to enable seamless interaction between Large Language Model (LLM) applications and external tools or resources. Within a short period, thousands of MCP services have already been developed and deployed. However, the client-server integration architecture inherent in MCP may expand the attack surface against LLM Agent systems, introducing new vulnerabilities that allow attackers to exploit by designing malicious MCP servers. In th…

AttenTrack: Mobile User Attention Awareness Based on Context and External Distractions
Yutong Lin, Suyuan Liu, Kaiwen Guo, Haohua Du, Chao Liu, Xiang-Yang Li
https://arxiv.org/abs/2509.01414

AttenTrack: Mobile User Attention Awareness Based on Context and External Distractions
In the mobile internet era, managing limited attention amid information overload is crucial for enhancing collaboration and information delivery. However, current attention-aware systems often depend on wearables or personalized data, limiting their scalability and cross-context adaptability. Inspired by psychological theories, we attempt to treat mobile notifications as naturally occurring external distractions and infer users' attention states based on their response behaviors and contextual …

Medical Data Pecking: A Context-Aware Approach for Automated Quality Evaluation of Structured Medical Data
Irena Girshovitz, Atai Ambus, Moni Shahar, Ran Gilad-Bachrach
https://arxiv.org/abs/2507.02628

Medical Data Pecking: A Context-Aware Approach for Automated Quality Evaluation of Structured Medical Data
Background: The use of Electronic Health Records (EHRs) for epidemiological studies and artificial intelligence (AI) training is increasing rapidly. The reliability of the results depends on the accuracy and completeness of EHR data. However, EHR data often contain significant quality issues, including misrepresentations of subpopulations, biases, and systematic errors, as they are primarily collected for clinical and billing purposes. Existing quality assessment methods remain insufficient, la…

The term "context engineering" is gaining traction over "prompt engineering" as it better describes the skill of providing LLMs with the necessary information (Simon Willison/Simon Willison's Weblog)
https://simonwillison.net/2025/Jun/27/context-…

Context-aware gate set tomography: Improving the self-consistent characterization of trapped-ion universal gate sets by leveraging non-Markovianity
Pablo Vi\~nas, Alejandro Bermudez
https://arxiv.org/abs/2507.02542

Context-aware gate set tomography: Improving the self-consistent characterization of trapped-ion universal gate sets by leveraging non-Markovianity
To progress in the characterization of noise for current quantum computers, gate set tomography (GST) has emerged as a self-consistent tomographic protocol that can accurately estimate the complete set of noisy quantum gates, state preparations, and measurements. In its original incarnation, GST improves the estimation precision by applying the gates sequentially, provided that the noise makes them a set of fixed completely-positive and trace preserving (CPTP) maps independent of the history of…

AdaptCache: KV Cache Native Storage Hierarchy for Low-Delay and High-Quality Language Model Serving
Shaoting Feng, Hanchen Li, Kuntai Du, Zhuohan Gu, Yuhan Liu, Jiayi Yao, Siddhant Ray, Samuel Shen, Yihua Cheng, Ganesh Ananthanarayanan, Junchen Jiang
https://arxiv.org/abs/2509.00105

AdaptCache: KV Cache Native Storage Hierarchy for Low-Delay and High-Quality Language Model Serving
Large language model (LLM) applications often reuse previously processed context, such as chat history and documents, which introduces significant redundant computation. Existing LLM serving systems address such redundant computation by storing the KV caches of processed context and loading the corresponding KV cache when a new request reuses the context. Further, as these LLM applications scale, the total size of KV caches becomes excessively large and requires both DRAM and SSD for full stora…

The president of El Salvador would not recognize Justice if it indicted him in The Hague and sentenced him for his crimes.
But let’s do it anyway just to be sure.
Context: https://apnews.com/article/kilmar-abrego-…

Totoro spoilers
Re the quoted post from @…:
The very first time I saw My Neighbor Totoro, it was because a friend cajoled me into attending the student anime club’s showing with absolutely no context whatsoever except “you •have• to see this movie, Paul.” I had no idea what genre it was. I had no idea that it was a movie considered suitable for kids. The last anime I’d seen was IIRC Ghost in the Shell; I was ready for anything.
When Mei went missing, I thought, “omg, is this a tragedy? I think this story is a tragedy!” The whole time they were looking for her, I was absolutely terrified. I thought for sure they’d found her sandal. I still tear up when they find her now, on every rewatching.
I’m so glad for that first viewing. It’s a much better movie that way. When you’re expecting an innocent movie about cute forest plushies, you see that. But when you see that it could be a tragedy — the mother’s shadowy illness, the lost child — it hits hard. https://wandering.shop/@Violinknitter/114792748619083661

Multi-Focus Probes for Context-Preserving Network Exploration and Interaction in Immersive Analytics
Eric Zimmermann, Stefan Bruckner
https://arxiv.org/abs/2507.01140

Multi-Focus Probes for Context-Preserving Network Exploration and Interaction in Immersive Analytics
Immersive visualization of network data enables users to physically navigate and interact with complex structures, but managing transitions between detailed local (egocentric) views and global (exocentric) overviews remains a major challenge. We present a multifocus probe technique for immersive environments that allows users to instantiate multiple egocentric subgraph views while maintaining persistent links to the global network context. Each probe acts as a portable local focus, enabling fin…

In-context Clustering-based Entity Resolution with Large Language Models: A Design Space Exploration
Jiajie Fu, Haitong Tang, Arijit Khan, Sharad Mehrotra, Xiangyu Ke, Yunjun Gao
https://arxiv.org/abs/2506.02509

In-context Clustering-based Entity Resolution with Large Language Models: A Design Space Exploration
Entity Resolution (ER) is a fundamental data quality improvement task that identifies and links records referring to the same real-world entity. Traditional ER approaches often rely on pairwise comparisons, which can be costly in terms of time and monetary resources, especially with large datasets. Recently, Large Language Models (LLMs) have shown promising results in ER tasks. However, existing methods typically focus on pairwise matching, missing the potential of LLMs to perform clustering di…

A Context-Aware Dual-Metric Framework for Confidence Estimation in Large Language Models
Mingruo Yuan, Shuyi Zhang, Ben Kao
https://arxiv.org/abs/2508.00600 https://

A Context-Aware Dual-Metric Framework for Confidence Estimation in Large Language Models
Accurate confidence estimation is essential for trustworthy large language models (LLMs) systems, as it empowers the user to determine when to trust outputs and enables reliable deployment in safety-critical applications. Current confidence estimation methods for LLMs neglect the relevance between responses and contextual information, a crucial factor in output quality evaluation, particularly in scenarios where background knowledge is provided. To bridge this gap, we propose CRUX (Context-awar…

Having legal protection insurance is always a good idea to, but IMHO it’s absolutely essential for professors due to the particular context in which we work. But I find that it’s hardly ever discussed.
An obvious example is the question of the limits of academic freedom, but many potential issues are way more trivial. ⇢
#AcademicChatter

Having legal protection insurance is always a good idea to, but IMHO it’s absolutely essential for professors due to the particular context in which we work. But I find that it’s hardly ever discussed.
An obvious example is the question of the limits of academic freedom, but many potential issues are way more trivial. ⇢
#AcademicChatter

Having legal protection insurance is always a good idea to, but IMHO it’s absolutely essential for professors due to the particular context in which we work. But I find that it’s hardly ever discussed.
An obvious example is the question of the limits of academic freedom, but many potential issues are way more trivial. ⇢
#AcademicChatter

Context dependent adaptation in a neural computation
Charles J. Edelson, Sima Setayeshgar, William Bialek, Rob R. de Ruyter van Steveninck
https://arxiv.org/abs/2509.01760 https…

Context dependent adaptation in a neural computation
Brains adapt to the statistical structure of their input. In the visual system, local light intensities change rapidly, the variance of the intensity changes more slowly, and the dynamic range of contrast itself changes more slowly still. We use a motion-sensitive neuron in the fly visual system to probe this hierarchy of adaptation phenomena, delivering naturalistic stimuli that have been simplified to have a clear separation of time scales. We show that the neural response to visual motion de…

Context-Aware Code Wiring Recommendation with LLM-based Agent
Taiming Wang, Yanjie Jiang, Chunhao Dong, Yuxia Zhang, Hui Liu
https://arxiv.org/abs/2507.01315

Context-Aware Code Wiring Recommendation with LLM-based Agent
Copy-paste-modify is a widespread and pragmatic practice in software development, where developers adapt reused code snippets, sourced from platforms such as Stack Overflow, GitHub, or LLM outputs, into their local codebase. A critical yet underexplored aspect of this adaptation is code wiring, which involves substituting unresolved variables in the pasted code with suitable ones from the surrounding context. Existing solutions either rely on heuristic rules or historical templates, often faili…

@… For those of you that manually installed 7.5RC1, these are the additional changes since that version:
• [Crash] On tab strip manipulation
• [Crash][Search][Sync] In search engines code, triggered by sync (VB-118299)
• [Crash][macOS] On adding an new account (VB-118293)
• [Menus][Tabs] The tab bar context menu is no longer accessible: pa…

A Dynamic Framework for Semantic Grouping of Common Data Elements (CDE) Using Embeddings and Clustering
Madan Krishnamurthy, Daniel Korn, Melissa A Haendel, Christopher J Mungall, Anne E Thessen
https://arxiv.org/abs/2506.02160

A Dynamic Framework for Semantic Grouping of Common Data Elements (CDE) Using Embeddings and Clustering
This research aims to develop a dynamic and scalable framework to facilitate harmonization of Common Data Elements (CDEs) across heterogeneous biomedical datasets by addressing challenges such as semantic heterogeneity, structural variability, and context dependence to streamline integration, enhance interoperability, and accelerate scientific discovery. Our methodology leverages Large Language Models (LLMs) for context-aware text embeddings that convert CDEs into dense vectors capturing semant…

Overlapping IPv4, IPv6, and TCP data: exploring errors, test case context and multiple overlaps inside network stacks and NIDSes with PYROLYSE
Lucas Aubard, Johan Mazel, Gilles Guette, Pierre Chifflier
https://arxiv.org/abs/2508.00735

Overlapping IPv4, IPv6, and TCP data: exploring errors, test case context and multiple overlaps inside network stacks and NIDSes with PYROLYSE
IP fragmentation and TCP segmentation allow for splitting large data packets into smaller ones, e.g., for transmission across network links of limited capacity. These mechanisms permit complete or partial overlaps with different data on the overlapping portions. IPv4, IPv6, and TCP reassembly policies, i.e., the data chunk preferences that depend on the overlap types, differ across protocol implementations. This leads to vulnerabilities, as NIDSes may interpret the packet differently from the m…

Trying to figure out why Gemini AI in Gmail is so very bad. It seems to not have access to all of my email and whatever RAG-like thing they are doing is very stupid.
Here's Gemini saying my oldest email matching some keywords in 2021 is from March 2021. If I explicitly tell it to look in February 2021, it finds it. There's only 11 emails that match in all of 2021, it's not like I'm blowing up the context window.

Combinatorics of monoidal actions in Lie-algebraic context
Volodymyr Mazorchuk, Xiaoyu Zhu
https://arxiv.org/abs/2509.01404 https://arxiv.org/pdf/2509.0140…

Combinatorics of monoidal actions in Lie-algebraic context
This paper is, essentially, a survey related to the problem of understanding the combinatorics of the action of the monoidal category of finite dimensional modules over a simple finite dimensional Lie algebra on various categories of Lie algebra modules. A special attention is payed to the Lie algebras $\mathfrak{sl}_2$ and $\mathfrak{sl}_3$. A few new general results are collected at the end.

Linear Attention with Global Context: A Multipole Attention Mechanism for Vision and Physics
Alex Colagrande, Paul Caillon, Eva Feillet, Alexandre Allauzen
https://arxiv.org/abs/2507.02748

Linear Attention with Global Context: A Multipole Attention Mechanism for Vision and Physics
Transformers have become the de facto standard for a wide range of tasks, from image classification to physics simulations. Despite their impressive performance, the quadratic complexity of standard Transformers in both memory and time with respect to the input length makes them impractical for processing high-resolution inputs. Therefore, several variants have been proposed, the most successful relying on patchification, downsampling, or coarsening techniques, often at the cost of losing the f…

Context-Driven Knowledge Graph Completion with Semantic-Aware Relational Message Passing
Siyuan Li, Ruitong Liu, Yan Wen, Te Sun
https://arxiv.org/abs/2506.23141

Context-Driven Knowledge Graph Completion with Semantic-Aware Relational Message Passing
Semantic context surrounding a triplet $(h, r, t)$ is crucial for Knowledge Graph Completion (KGC), providing vital cues for prediction. However, traditional node-based message passing mechanisms, when applied to knowledge graphs, often introduce noise and suffer from information dilution or over-smoothing by indiscriminately aggregating information from all neighboring edges. To address this challenge, we propose a semantic-aware relational message passing. A core innovation of this framework …

A post from the archive 📫:
Building debugging context for Copilot Chat
https://www.poppastring.com/blog/building-debugging-context-for-copilot-chat

„My new system is, simply, no system at all. I write what I think. I delete what I don’t need. I don’t capture everything. I don’t try to. I read what I feel like. I think in conversation, in movement, in context. I don’t build a second brain. I inhabit the first.”
The PKM-movement never took off for me since it feels like building and bulding and building… Just one more note! It’s like productivity snake-oil.

I Deleted My Second Brain
Why I Erased 10,000 Notes, 7 Years of Ideas, and Every Thought I Tried to Save

Exposing and Mitigating Calibration Biases and Demographic Unfairness in MLLM Few-Shot In-Context Learning for Medical Image Classification
Xing Shen, Justin Szeto, Mingyang Li, Hengguan Huang, Tal Arbel
https://arxiv.org/abs/2506.23298

Exposing and Mitigating Calibration Biases and Demographic Unfairness in MLLM Few-Shot In-Context Learning for Medical Image Classification
Multimodal large language models (MLLMs) have enormous potential to perform few-shot in-context learning in the context of medical image analysis. However, safe deployment of these models into real-world clinical practice requires an in-depth analysis of the accuracies of their predictions, and their associated calibration errors, particularly across different demographic subgroups. In this work, we present the first investigation into the calibration biases and demographic unfairness of MLLMs'…

Analysing Temporal Reasoning in Description Logics Using Formal Grammars
Camille Bourgaux, Anton Gnatenko, Micha\"el Thomazo
https://arxiv.org/abs/2508.00575 https://

Analysing Temporal Reasoning in Description Logics Using Formal Grammars
We establish a correspondence between (fragments of) $\mathcal{TEL}^\bigcirc$, a temporal extension of the $\mathcal{EL}$ description logic with the LTL operator $\bigcirc^k$, and some specific kinds of formal grammars, in particular, conjunctive grammars (context-free grammars equipped with the operation of intersection). This connection implies that $\mathcal{TEL}^\bigcirc$ does not possess the property of ultimate periodicity of models, and further leads to undecidability of query answering …

Chrome (very good developers, some of the best, have not made any UX mistakes in the past 5 minutes) gives me a duplicate "Open in reading mode" context menu entry for some reason

This https://arxiv.org/abs/2505.04316 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_hept…

On local fields invariant under the action of topological defects
In the context of rational conformal field theories (RCFT) we look into the problem of constructing and classifying pairs consisting of a local operator and a topological defect which commutes or anticommutes with it. We discuss the bulk and boundary versions of the problem. In the latter one considers a conformal boundary condition, a boundary operator on it and a junction with a topological defect. In the case of the charge conjugation modular invariant commuting configurations in each proble…

Homotopy equivalence of digital pictures in $\mathbb{Z}^2$
Dae-Woong Lee, P. Christopher Staecker
https://arxiv.org/abs/2509.03023 https://arxiv.org/pdf/25…

Homotopy equivalence of digital pictures in $\mathbb{Z}^2$
We investigate the properties of digital homotopy in the context of digital pictures $(X,κ,\bar κ)$, where $X\subsetneq \mathbb{Z}^n$ is a finite set, $κ$ is an adjacency relation on $X$, and $\bar κ$ is an adjacency relation on the complement of $X$. In particular we focus on homotopy equivalence between digital pictures in $\mathbb{Z}^2$. We define a numerical homotopy-type invariant for digital pictures in $\mathbb{Z}^2$ called the outer perimeter, which is a basic tool for distinguishin…

KVComp: A High-Performance, LLM-Aware, Lossy Compression Framework for KV Cache
Bo Jiang, Taolue Yang, Youyuan Liu, Chengming Zhang, Xubin He, Sian Jin
https://arxiv.org/abs/2509.00579

KVComp: A High-Performance, LLM-Aware, Lossy Compression Framework for KV Cache
Transformer-based large language models (LLMs) demonstrate impressive potential in various practical applications. However, long context inference poses a significant challenge due to the enormous memory requirements of the key-value (KV) cache, which can scale to multiple gigabytes as sequence length and batch size increase. In this paper, we present KVComp, a generic and efficient KV cache management framework optimized for long-text generation that synergistically works with both latency-cri…

New orientable sequences
Chris J Mitchell, Peter R Wild
https://arxiv.org/abs/2507.02526 https://arxiv.org/pdf/2507.02526

New orientable sequences
Orientable sequences of order n are infinite periodic sequences with symbols drawn from a finite alphabet of size k with the property that any particular subsequence of length n occurs at most once in a period in either direction. They were introduced in the early 1990s in the context of possible applications in position sensing. Bounds on the period of such sequences and a range of methods of construction have been devised, although apart from very small cases a significant gap remains between…

💰 Supports multiple dimensions and quantization options - binary 512d version outperforms OpenAI-v3-large while reducing vector database costs by 99.48%
🔍 Processes entire documents in single pass to generate chunk embeddings enriched with document-level context
🎯 Less sensitive to chunking strategies compared to traditional context-agnostic embedding models

IC-Custom: Diverse Image Customization via In-Context Learning
Yaowei Li, Xiaoyu Li, Zhaoyang Zhang, Yuxuan Bian, Gan Liu, Xinyuan Li, Jiale Xu, Wenbo Hu, Yating Liu, Lingen Li, Jing Cai, Yuexian Zou, Yancheng He, Ying Shan
https://arxiv.org/abs/2507.01926

IC-Custom: Diverse Image Customization via In-Context Learning
Image customization, a crucial technique for industrial media production, aims to generate content that is consistent with reference images. However, current approaches conventionally separate image customization into position-aware and position-free customization paradigms and lack a universal framework for diverse customization, limiting their applications across various scenarios. To overcome these limitations, we propose IC-Custom, a unified framework that seamlessly integrates position-awa…

PIMMS: Pulsation-Informed Magnetic Mapping of Stars with Zeeman-Doppler Imaging I. Formalism and numerical tests
C. Gutteridge, C. Neiner, C. Catala, C. P. Folsom
https://arxiv.org/abs/2509.02152

PIMMS: Pulsation-Informed Magnetic Mapping of Stars with Zeeman-Doppler Imaging I. Formalism and numerical tests
Context. Magneto-asteroseismology is a novel technique allowing for more precise determinations of internal properties of magnetic pulsating stars, but requires an accurate characterisation of the surface magnetic field, not previously possible with Zeeman-Doppler Imaging (ZDI) due to the time-dependent surface velocity of pulsating stars. Aims. We aim to develop a new version of ZDI, which creates an additional surface velocity map, that includes the time-dependent velocities of surface elemen…

StructCoh: Structured Contrastive Learning for Context-Aware Text Semantic Matching
Chao Xue, Ziyuan Gao
https://arxiv.org/abs/2509.02033 https://arxiv.org…

StructCoh: Structured Contrastive Learning for Context-Aware Text Semantic Matching
Text semantic matching requires nuanced understanding of both structural relationships and fine-grained semantic distinctions. While pre-trained language models excel at capturing token-level interactions, they often overlook hierarchical structural patterns and struggle with subtle semantic discrimination. In this paper, we proposed StructCoh, a graph-enhanced contrastive learning framework that synergistically combines structural reasoning with representation space optimization. Our approach …

Towards Human-like Preference Profiling in Sequential Recommendation
Zhongyu Ouyang, Qianlong Wen, Chunhui Zhang, Yanfang Ye, Soroush Vosoughi
https://arxiv.org/abs/2506.02261

Towards Human-like Preference Profiling in Sequential Recommendation
Sequential recommendation systems aspire to profile users by interpreting their interaction histories, echoing how humans make decisions by weighing experience, relative preference strength, and situational relevance. Yet, existing large language model (LLM)-based recommenders often fall short of mimicking the flexible, context-aware decision strategies humans exhibit, neglecting the structured, dynamic, and context-aware mechanisms fundamental to human behaviors. To bridge this gap, we propose…

SleepWalk: Exploiting Context Switching and Residual Power for Physical Side-Channel Attacks
Sahan Sanjaya, Aruna Jayasena, Prabhat Mishra
https://arxiv.org/abs/2507.22306 https…

SleepWalk: Exploiting Context Switching and Residual Power for Physical Side-Channel Attacks
Context switching is utilized by operating systems to change the execution context between application programs. It involves saving and restoring the states of multiple registers and performing a pipeline flush to remove any pre-fetched instructions, leading to a higher instantaneous power consumption compared to typical program execution. In this paper, we introduce a physical power side-channel leakage source that exploits the power spike observed during a context switch, triggered by the inb…

Visual Contextual Attack: Jailbreaking MLLMs with Image-Driven Context Injection
Ziqi Miao, Yi Ding, Lijun Li, Jing Shao
https://arxiv.org/abs/2507.02844 h…

Visual Contextual Attack: Jailbreaking MLLMs with Image-Driven Context Injection
With the emergence of strong visual-language capabilities, multimodal large language models (MLLMs) have demonstrated tremendous potential for real-world applications. However, the security vulnerabilities exhibited by the visual modality pose significant challenges to deploying such models in open-world environments. Recent studies have successfully induced harmful responses from target MLLMs by encoding harmful textual semantics directly into visual inputs. However, in these approaches, the v…

Exploring the Design Space of Fair Tree Learning Algorithms
Kiara Stempel, Mattia Cerrato, Stefan Kramer
https://arxiv.org/abs/2509.03204 https://arxiv.org…

Exploring the Design Space of Fair Tree Learning Algorithms
Decision trees have been studied extensively in the context of fairness, aiming to maximize prediction performance while ensuring non-discrimination against different groups. Techniques in this space usually focus on imposing constraints at training time, constraining the search space so that solutions which display unacceptable values of relevant metrics are not considered, discarded, or discouraged. If we assume one target variable y and one sensitive attribute s, the design space of tree lea…

CooT: Learning to Coordinate In-Context with Coordination Transformers
Huai-Chih Wang, Hsiang-Chun Chuang, Hsi-Chun Cheng, Dai-Jie Wu, Shao-Hua Sun
https://arxiv.org/abs/2506.23549

CooT: Learning to Coordinate In-Context with Coordination Transformers
Effective coordination among artificial agents in dynamic and uncertain environments remains a significant challenge in multi-agent systems. Existing approaches, such as self-play and population-based methods, either generalize poorly to unseen partners or require extensive training. To overcome these limitations, we propose Coordination Transformers (CooT), a novel in-context coordination framework that uses recent interaction histories to adapt to unseen partners rapidly. Unlike previous appr…

Dynamically Adaptive Reasoning via LLM-Guided MCTS for Efficient and Context-Aware KGQA
Yingxu Wang, Shiqi Fan, Mengzhu Wang, Siwei Liu
https://arxiv.org/abs/2508.00719 https://…

Dynamically Adaptive Reasoning via LLM-Guided MCTS for Efficient and Context-Aware KGQA
Knowledge Graph Question Answering (KGQA) aims to interpret natural language queries and perform structured reasoning over knowledge graphs by leveraging their relational and semantic structures to retrieve accurate answers. Recent KGQA methods primarily follow either retrieve-then-reason paradigm, relying on GNNs or heuristic rules for static paths extraction, or dynamic path generation strategies that use large language models (LLMs) with prompting to jointly perform retrieval and reasoning. …

Effective Explanations for Belief-Desire-Intention Robots: When and What to Explain
Cong Wang, Roberto Calandra, Verena Kl\"os
https://arxiv.org/abs/2507.02016

Effective Explanations for Belief-Desire-Intention Robots: When and What to Explain
When robots perform complex and context-dependent tasks in our daily lives, deviations from expectations can confuse users. Explanations of the robot's reasoning process can help users to understand the robot intentions. However, when to provide explanations and what they contain are important to avoid user annoyance. We have investigated user preferences for explanation demand and content for a robot that helps with daily cleaning tasks in a kitchen. Our results show that users want explanatio…

Context, Credibility, and Control: User Reflections on AI Assisted Misinformation Tools
Varun Sangwan, Heidi Makitalo
https://arxiv.org/abs/2506.22940 http…

Context, Credibility, and Control: User Reflections on AI Assisted Misinformation Tools
This paper investigates how collaborative AI systems can enhance user agency in identifying and evaluating misinformation on social media platforms. Traditional methods, such as personal judgment or basic fact-checking, often fall short when faced with emotionally charged or context-deficient content. To address this, we designed and evaluated an interactive interface that integrates collaborative AI features, including real-time explanations, source aggregation, and debate-style interaction. T…

LCQNN: Linear Combination of Quantum Neural Networks
Hongshun Yao, Xia Liu, Mingrui Jing, Xin Wang
https://arxiv.org/abs/2507.02832 https://

LCQNN: Linear Combination of Quantum Neural Networks
Quantum neural networks combine quantum computing with advanced data-driven methods, offering promising applications in quantum machine learning. However, the optimal paradigm of quantum neural networks in the context trainability and expressivity remains an open question. To overcome this issue, we first introduce a framework called Linear Combination of Quantum Neural Networks (LCQNN), which leverages the linear combination of unitaries concept to coordinate the composition of QNNs, thereby b…

Bug Fixing with Broader Context: Enhancing LLM-Based Program Repair via Layered Knowledge Injection
Ramtin Ehsani, Esteban Parra, Sonia Haiduc, Preetha Chatterjee
https://arxiv.org/abs/2506.24015

Bug Fixing with Broader Context: Enhancing LLM-Based Program Repair via Layered Knowledge Injection
Prompting LLMs with bug-related context (e.g., error messages, stack traces) improves automated program repair, but many bugs still remain unresolved. In real-world projects, developers often rely on broader repository and project-level context beyond the local code to resolve such bugs. In this paper, we investigate how automatically extracting and providing such knowledge can improve LLM-based program repair. We propose a layered knowledge injection framework that incrementally augments LLMs …

Soft Theorems in Chern-Simons Matter Theories
Avi Wadhwa
https://arxiv.org/abs/2509.02542 https://arxiv.org/pdf/2509.02542…

Soft Theorems in Chern-Simons Matter Theories
I will describe tree level soft theorems for Chern Simons QED and QCD in 4+1 dimensions. Soft theorems have been studied in various context in the past. Recently, there has been renewed interest in these theorems due to their connection with asymptotic symmetries and gauge invariance. Chern Simons theories are gauge invariant only up to a boundary term. Hence, they make for an interesting case to study the relationship between gauge invariance and soft theorems. The aim of this work is to inclu…

A Two-Stage Strategy for Mitosis Detection Using Improved YOLO11x Proposals and ConvNeXt Classification
Jie Xiao, Mengye Lyu, Shaojun Liu
https://arxiv.org/abs/2509.02627 https:…

A Two-Stage Strategy for Mitosis Detection Using Improved YOLO11x Proposals and ConvNeXt Classification
MIDOG 2025 Track 1 requires mitosis detection in whole-slide images (WSIs) containing non-tumor, inflamed, and necrotic regions. Due to the complicated and heterogeneous context, as well as possible artifacts, there are often false positives and false negatives, thus degrading the detection F1-score. To address this problem, we propose a two-stage framework. Firstly, an improved YOLO11x, integrated with EMA attention and LSConv, is employed to generate mitosis candidates. We use a low confidenc…

Understanding and Improving Length Generalization in Recurrent Models
Ricardo Buitrago Ruiz, Albert Gu
https://arxiv.org/abs/2507.02782 https://

Understanding and Improving Length Generalization in Recurrent Models
Recently, recurrent models such as state space models and linear attention have become popular due to their linear complexity in the sequence length. Thanks to their recurrent nature, in principle they can process arbitrarily long sequences, but their performance sometimes drops considerably beyond their training context lengths-i.e. they fail to length generalize. In this work, we provide comprehensive empirical and theoretical analysis to support the unexplored states hypothesis, which posits…

In-context learning for the classification of manipulation techniques in phishing emails
Antony Dalmiere (LAAS-TRUST, LAAS), Guillaume Auriol (LAAS-TRUST, INSA Toulouse), Vincent Nicomette (LAAS-TSF, LAAS), Pascal Marchand (LERASS)
https://arxiv.org/abs/2506.22515

In-context learning for the classification of manipulation techniques in phishing emails
Traditional phishing detection often overlooks psychological manipulation. This study investigates using Large Language Model (LLM) In-Context Learning (ICL) for fine-grained classification of phishing emails based on a taxonomy of 40 manipulation techniques. Using few-shot examples with GPT-4o-mini on real-world French phishing emails (SignalSpam), we evaluated performance against a human-annotated test set (100 emails). The approach effectively identifies prevalent techniques (e.g., Baiting, …

Geometric Visual Servo Via Optimal Transport
Ethan Canzini, Simon Pope, Ashutosh Tiwari
https://arxiv.org/abs/2506.02768 https://arxi…

Geometric Visual Servo Via Optimal Transport
When developing control laws for robotic systems, the principle factor when examining their performance is choosing inputs that allow smooth tracking to a reference input. In the context of robotic manipulation, this involves translating an object or end-effector from an initial pose to a target pose. Robotic manipulation control laws frequently use vision systems as an error generator to track features and produce control inputs. However, current control algorithms don't take into account the …

Quantum Key-Recovery Attacks on FBC Algorithm
Yan-Ying Zhu, Bin-Bin Cai, Fei Gao, Song Lin
https://arxiv.org/abs/2508.00448 https://arxiv.org/pdf/2508.0044…

Quantum Key-Recovery Attacks on FBC Algorithm
With the advancement of quantum computing, symmetric cryptography faces new challenges from quantum attacks. These attacks are typically classified into two models: Q1 (classical queries) and Q2 (quantum superposition queries). In this context, we present a comprehensive security analysis of the FBC algorithm considering quantum adversaries with different query capabilities. In the Q2 model, we first design 4-round polynomial-time quantum distinguishers for FBC-F and FBC-KF structures, and then…

Finding My Way: Influence of Different Audio Augmented Reality Navigation Cues on User Experience and Subjective Usefulness
Sina Hinzmann, Francesco Vona, Juliane Henning, Mohamed Amer, Omar Abdellatif, Tanja Kojic, Jan-Niklas Voigt-Antons
https://arxiv.org/abs/2509.03199

Finding My Way: Influence of Different Audio Augmented Reality Navigation Cues on User Experience and Subjective Usefulness
As augmented reality (AR) becomes increasingly prevalent in mobile and context-aware applications, the role of auditory cues in guiding users through physical environments is becoming critical. This study investigates the effectiveness and user experience of various categories of audio cues, including fully non-verbal sounds and speech-derived Spearcons, during outdoor navigation tasks using the Meta Quest 3 headset. Twenty participants navigated five outdoor routes using audio-only cue types: …

Computing $c$- and $a$-functions from entanglement
Konstantinos Boutivas, Dimitrios Katsinis, Georgios Pastras, Nikolaos Tetradis
https://arxiv.org/abs/2509.03259 https://

Computing $c$- and $a$-functions from entanglement
We confirm the direct connection between entanglement entropy and the notion of irreversibility in the renormalization-group flow in the context of a simple theory for which a calculation from first principles is feasible. The change of the entanglement entropy for a spherical entangling surface as its radius grows from zero to infinity corresponds to the flow from the UV to the IR. Through analytical and numerical means, we compute the entanglement entropy for a free massive scalar theory, mak…

A Large-Scale Evolvable Dataset for Model Context Protocol Ecosystem and Security Analysis
Zhiwei Lin, Bonan Ruan, Jiahao Liu, Weibo Zhao
https://arxiv.org/abs/2506.23474

A Large-Scale Evolvable Dataset for Model Context Protocol Ecosystem and Security Analysis
The Model Context Protocol (MCP) has recently emerged as a standardized interface for connecting language models with external tools and data. As the ecosystem rapidly expands, the lack of a structured, comprehensive view of existing MCP artifacts presents challenges for research. To bridge this gap, we introduce MCPCorpus, a large-scale dataset containing around 14K MCP servers and 300 MCP clients. Each artifact is annotated with 20+ normalized attributes capturing its identity, interface conf…

Metamorphic Testing of Multimodal Human Trajectory Prediction
Helge Spieker, Nadjib Lazaar, Arnaud Gotlieb, Nassim Belmecheri
https://arxiv.org/abs/2509.01294 https://

Metamorphic Testing of Multimodal Human Trajectory Prediction
Context: Predicting human trajectories is crucial for the safety and reliability of autonomous systems, such as automated vehicles and mobile robots. However, rigorously testing the underlying multimodal Human Trajectory Prediction (HTP) models, which typically use multiple input sources (e.g., trajectory history and environment maps) and produce stochastic outputs (multiple possible future paths), presents significant challenges. The primary difficulty lies in the absence of a definitive test …

This https://arxiv.org/abs/2505.07802 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csRO_…

Improving Trajectory Stitching with Flow Models
Generative models have shown great promise as trajectory planners, given their affinity to modeling complex distributions and guidable inference process. Previous works have successfully applied these in the context of robotic manipulation but perform poorly when the required solution does not exist as a complete trajectory within the training set. We identify that this is a result of being unable to plan via stitching, and subsequently address the architectural and dataset choices needed to re…

The Role of the Satellite in Quantum Information Networks
Luca Paccard, Valentin Leloup, Luca Lazzarini, Agathe Blaise, Mailys Guerault, Mickael Faugeron, Fabrice Arnal, Mathieu Bertrand, Raphael Aymeric, Michel Sotom, St\'ephanie Molin, Patrick G\'elard, Pierre Besancenot, Cyrille Laborde, Laurent de Forges de Parny, Mathias van den Bossche
https://

The Role of the Satellite in Quantum Information Networks
Quantum Information Networks (QIN) attract increasing interest, as they will enable interconnection of multiple quantum devices in a distributed organization thus enhancing intrinsic computing, sensing, and security capabilities. The core mechanism of a QIN is quantum state swapping, based on teleportation, which consumes quantum entanglement, and which can be seen in this context as a new kind of network resource. The satellite is expected to play a central role for supporting global connectiv…

Exploring listeners' perceptions of AI-generated and human-composed music for functional emotional applications
Kimaya Lecamwasam, Tishya Ray Chaudhuri
https://arxiv.org/abs/2506.02856

Exploring listeners' perceptions of AI-generated and human-composed music for functional emotional applications
This work investigates how listeners perceive and evaluate AI-generated as compared to human-composed music in the context of emotional resonance and regulation. Across a mixed-methods design, participants were exposed to both AI and human music under various labeling conditions (music correctly labeled as AI- or human-origin, music incorrectly labeled as AI- or human-origin, and unlabeled music) and emotion cases (Calm and Upbeat), and were asked to rate preference, efficacy of target emotion …

Leveraging In-Context Learning for Political Bias Testing of LLMs
Patrick Haller, Jannis Vamvas, Rico Sennrich, Lena A. J\"ager
https://arxiv.org/abs/2506.22232

Leveraging In-Context Learning for Political Bias Testing of LLMs
A growing body of work has been querying LLMs with political questions to evaluate their potential biases. However, this probing method has limited stability, making comparisons between models unreliable. In this paper, we argue that LLMs need more context. We propose a new probing task, Questionnaire Modeling (QM), that uses human survey data as in-context examples. We show that QM improves the stability of question-based bias evaluation, and demonstrate that it may be used to compare instruct…

Managing Power Gaps as a Topic of Pair Programming Skill: A Grounded Theory
Linus Ververs, Lutz Prechelt
https://arxiv.org/abs/2508.00462 https://arxiv.org…

Managing Power Gaps as a Topic of Pair Programming Skill: A Grounded Theory
Context: Pair Programming as a work mode is used (occasionally or frequently) throughout professional software development. Objective: Understand what power-related phenomena occur in pair programming as it is used in industry; give advice to practitioners on how to do better pair programming. Method: Analyze 22 industrial pair programming sessions using Grounded Theory Methodology. Formulate a Grounded Theory on power-related behaviors. Run a survey with 292 participants about that theory. Use…

From Long Videos to Engaging Clips: A Human-Inspired Video Editing Framework with Multimodal Narrative Understanding
Xiangfeng Wang, Xiao Li, Yadong Wei, Xueyu Song, Yang Song, Xiaoqiang Xia, Fangrui Zeng, Zaiyi Chen, Liu Liu, Gu Xu, Tong Xu
https://arxiv.org/abs/2507.02790

From Long Videos to Engaging Clips: A Human-Inspired Video Editing Framework with Multimodal Narrative Understanding
The rapid growth of online video content, especially on short video platforms, has created a growing demand for efficient video editing techniques that can condense long-form videos into concise and engaging clips. Existing automatic editing methods predominantly rely on textual cues from ASR transcripts and end-to-end segment selection, often neglecting the rich visual context and leading to incoherent outputs. In this paper, we propose a human-inspired automatic video editing framework (HIVE)…

EIM-TRNG: Obfuscating Deep Neural Network Weights with Encoding-in-Memory True Random Number Generator via RowHammer
Ranyang Zhou, Abeer Matar A. Almalky, Gamana Aragonda, Sabbir Ahmed, Filip Roth Tr{\o}nnes-Christensen, Adnan Siraj Rakin, Shaahin Angizi
https://arxiv.org/abs/2507.02206 …

EIM-TRNG: Obfuscating Deep Neural Network Weights with Encoding-in-Memory True Random Number Generator via RowHammer
True Random Number Generators (TRNGs) play a fundamental role in hardware security, cryptographic systems, and data protection. In the context of Deep NeuralNetworks (DNNs), safeguarding model parameters, particularly weights, is critical to ensure the integrity, privacy, and intel-lectual property of AI systems. While software-based pseudo-random number generators are widely used, they lack the unpredictability and resilience offered by hardware-based TRNGs. In this work, we propose a novel an…

Path Planning using a One-shot-sampling Skeleton Map
Gabriel O. Flores-Aquino, Octavio Gutierrez-Frias, Juan Irving Vasquez
https://arxiv.org/abs/2507.02328

Path Planning using a One-shot-sampling Skeleton Map
Path planning algorithms aim to compute a collision-free path, and many works focus on finding the optimal distance path. However, for some applications, a more suitable approach is to balance response time, safety of the paths, and path length. In this context, a skeleton map is a useful tool in graph-based schemes, as it provides an intrinsic representation of free configuration space. However, skeletonization algorithms are very resource-intensive, being primarily oriented towards image proc…

NeedleChain: Measuring Intact Long-Context Reasoning Capability of Large Language Models
Hyeonseok Moon, Heuiseok Lim
https://arxiv.org/abs/2507.22411 https://

NeedleChain: Measuring Intact Long-Context Reasoning Capability of Large Language Models
The Needle-in-a-Haystack (NIAH) benchmark is widely used to evaluate Large Language Models' (LLMs) ability to understand long contexts (LC). It evaluates the capability to identify query-relevant context within extensive query-irrelevant passages. Although this method serves as a widely accepted standard for evaluating long-context understanding, our findings suggest it may overestimate the true LC capability of LLMs. We demonstrate that even state-of-the-art models such as GPT-4o struggle to i…

SmartPoser: Arm Pose Estimation with a Smartphone and Smartwatch Using UWB and IMU Data
Nathan DeVrio, Vimal Mollyn, Chris Harrison
https://arxiv.org/abs/2509.03451 https://

SmartPoser: Arm Pose Estimation with a Smartphone and Smartwatch Using UWB and IMU Data
The ability to track a user's arm pose could be valuable in a wide range of applications, including fitness, rehabilitation, augmented reality input, life logging, and context-aware assistants. Unfortunately, this capability is not readily available to consumers. Systems either require cameras, which carry privacy issues, or utilize multiple worn IMUs or markers. In this work, we describe how an off-the-shelf smartphone and smartwatch can work together to accurately estimate arm pose. Moving be…

Dynamic Symbolic Execution for Semantic Difference Analysis of Component and Connector Architectures
Johanna Grahl, Bernhard Rumpe, Max Stachon, Sebastian St\"uber
https://arxiv.org/abs/2508.00749

Dynamic Symbolic Execution for Semantic Difference Analysis of Component and Connector Architectures
In the context of model-driven development, ensuring the correctness and consistency of evolving models is paramount. This paper investigates the application of Dynamic Symbolic Execution (DSE) for semantic difference analysis of component-and-connector architectures, specifically utilizing MontiArc models. We have enhanced the existing MontiArc-to-Java generator to gather both symbolic and concrete execution data at runtime, encompassing transition conditions, visited states, and internal vari…

chDzDT: Word-level morphology-aware language model for Algerian social media text
Abdelkrime Aries
https://arxiv.org/abs/2509.01772 https://arxiv.org/pdf/2…

chDzDT: Word-level morphology-aware language model for Algerian social media text
Pre-trained language models (PLMs) have substantially advanced natural language processing by providing context-sensitive text representations. However, the Algerian dialect remains under-represented, with few dedicated models available. Processing this dialect is challenging due to its complex morphology, frequent code-switching, multiple scripts, and strong lexical influences from other languages. These characteristics complicate tokenization and reduce the effectiveness of conventional word-…

Cryptographic Challenges: Masking Sensitive Data in Cyber Crimes through ASCII Art
Andres Alejandre, Kassandra Delfin, Victor Castano
https://arxiv.org/abs/2509.00059 https://…

Cryptographic Challenges: Masking Sensitive Data in Cyber Crimes through ASCII Art
The use of ASCII art as a novel approach to masking sensitive information in cybercrime, focusing on its potential role in protecting personal data during the delivery process and beyond, is presented. By examining the unique properties of ASCII art and its historical context, this study discusses the advantages and limitations of employing this technique in various cybercrime scenarios. Additionally, providing recommendations for enhancing data security practices and fostering a culture of pri…

Look: AI at Work! - Analysing Key Aspects of AI-support at the Work Place
Stefan Schiffer, Anna Milena Rothermel, Alexander Ferrein, Astrid Rosenthal-von der P\"utten
https://arxiv.org/abs/2509.02274

Look: AI at Work! - Analysing Key Aspects of AI-support at the Work Place
In this paper we present an analysis of technological and psychological factors of applying artificial intelligence (AI) at the work place. We do so for a number of twelve application cases in the context of a project where AI is integrated at work places and in work systems of the future. From a technological point of view we mainly look at the areas of AI that the applications are concerned with. This allows to formulate recommendations in terms of what to look at in developing an AI applicat…

Enhancing Speech Instruction Understanding and Disambiguation in Robotics via Speech Prosody
David Sasu, Kweku Andoh Yamoah, Benedict Quartey, Natalie Schluter
https://arxiv.org/abs/2506.02057

Enhancing Speech Instruction Understanding and Disambiguation in Robotics via Speech Prosody
Enabling robots to accurately interpret and execute spoken language instructions is essential for effective human-robot collaboration. Traditional methods rely on speech recognition to transcribe speech into text, often discarding crucial prosodic cues needed for disambiguating intent. We propose a novel approach that directly leverages speech prosody to infer and resolve instruction intent. Predicted intents are integrated into large language models via in-context learning to disambiguate and …

The Complexity Trap: Simple Observation Masking Is as Efficient as LLM Summarization for Agent Context Management
Tobias Lindenbauer, Igor Slinko, Ludwig Felder, Egor Bogomolov, Yaroslav Zharov
https://arxiv.org/abs/2508.21433

The Complexity Trap: Simple Observation Masking Is as Efficient as LLM Summarization for Agent Context Management
Large Language Model (LLM)-based agents solve complex tasks through iterative reasoning, exploration, and tool-use, a process that can result in long, expensive context histories. While state-of-the-art Software Engineering ( SE) agents like OpenHands or Cursor use LLM-based summarization to tackle this issue, it is unclear whether the increased complexity offers tangible performance benefits compared to simply omitting older observations. We present a systematic comparison of these strategies …

Understanding and Leveraging the Expert Specialization of Context Faithfulness in Mixture-of-Experts LLMs
Jun Bai, Minghao Tong, Yang Liu, Zixia Jia, Zilong Zheng
https://arxiv.org/abs/2508.19594

Understanding and Leveraging the Expert Specialization of Context Faithfulness in Mixture-of-Experts LLMs
Context faithfulness is essential for reliable reasoning in context-dependent scenarios. However, large language models often struggle to ground their outputs in the provided context, resulting in irrelevant responses. Inspired by the emergent expert specialization observed in mixture-of-experts architectures, this work investigates whether certain experts exhibit specialization in context utilization, offering a potential pathway toward targeted optimization for improved context faithfulness. …

E-PhishGen: Unlocking Novel Research in Phishing Email Detection
Luca Pajola, Eugenio Caripoti, Simeone Pizzi, Mauro Conti, Stefan Banzer, Giovanni Apruzzese
https://arxiv.org/abs/2509.01791

E-PhishGen: Unlocking Novel Research in Phishing Email Detection
Every day, our inboxes are flooded with unsolicited emails, ranging between annoying spam to more subtle phishing scams. Unfortunately, despite abundant prior efforts proposing solutions achieving near-perfect accuracy, the reality is that countering malicious emails still remains an unsolved dilemma. This "open problem" paper carries out a critical assessment of scientific works in the context of phishing email detection. First, we focus on the benchmark datasets that have been used to asses…

Pull Requests From The Classroom: Co-Developing Curriculum And Code
Dennis Zyska, Ilia Kuznetsov, Florian M\"uller, Iryna Gurevych
https://arxiv.org/abs/2508.00646 https://…

Pull Requests From The Classroom: Co-Developing Curriculum And Code
Educational technologies often misalign with instructors' pedagogical goals, forcing adaptations that compromise teaching efficacy. In this paper, we present a case study on the co-development of curriculum and technology in the context of a university course on scientific writing. Specifically, we examine how a custom-built peer feedback system was iteratively developed alongside the course to support annotation, feedback exchange, and revision. Results show that while co-development fostered …

Beyond Fixed: Variable-Length Denoising for Diffusion Large Language Models
Jinsong Li, Xiaoyi Dong, Yuhang Zang, Yuhang Cao, Jiaqi Wang, Dahua Lin
https://arxiv.org/abs/2508.00819

Beyond Fixed: Variable-Length Denoising for Diffusion Large Language Models
Diffusion Large Language Models (DLLMs) are emerging as a powerful alternative to the dominant Autoregressive Large Language Models, offering efficient parallel generation and capable global context modeling. However, the practical application of DLLMs is hindered by a critical architectural constraint: the need for a statically predefined generation length. This static length allocation leads to a problematic trade-off: insufficient lengths cripple performance on complex tasks, while excessive…

Benchmarking and Studying the LLM-based Code Review
Zhengran Zeng, Ruikai Shi, Keke Han, Yixin Li, Kaicheng Sun, Yidong Wang, Zhuohao Yu, Rui Xie, Wei Ye, Shikun Zhang
https://arxiv.org/abs/2509.01494 …

Benchmarking and Studying the LLM-based Code Review
Automated Code Review (ACR) is crucial for software quality, yet existing benchmarks often fail to reflect real-world complexities, hindering the evaluation of modern Large Language Models (LLMs). Current benchmarks frequently focus on fine-grained code units, lack complete project context, and use inadequate evaluation metrics. To address these limitations, we introduce SWRBench , a new benchmark comprising 1000 manually verified Pull Requests (PRs) from GitHub, offering PR-centric review with…

LeakSealer: A Semisupervised Defense for LLMs Against Prompt Injection and Leakage Attacks
Francesco Panebianco, Stefano Bonfanti, Francesco Trov\`o, Michele Carminati
https://arxiv.org/abs/2508.00602 …

LeakSealer: A Semisupervised Defense for LLMs Against Prompt Injection and Leakage Attacks
The generalization capabilities of Large Language Models (LLMs) have led to their widespread deployment across various applications. However, this increased adoption has introduced several security threats, notably in the forms of jailbreaking and data leakage attacks. Additionally, Retrieval Augmented Generation (RAG), while enhancing context-awareness in LLM responses, has inadvertently introduced vulnerabilities that can result in the leakage of sensitive information. Our contributions are t…

Unpacking Personal(?!) Health Informatics: An Investigation of Awareness, Understanding, And Leveraged Utility in India
Shyama Sastha Krishnamoorthy Srinivasan, Mohan Kumar, Pushpendra Singh
https://arxiv.org/abs/2509.01231

Unpacking Personal(?!) Health Informatics: An Investigation of Awareness, Understanding, And Leveraged Utility in India
Personal Health Informatics (PHI), which leverages digital tools and information systems to support health assessment and self-care, holds promise for empowering individuals and transforming healthcare delivery. However, barriers to its adoption remain underexplored in the Indian context. This study investigates PHI adoption among Indian users and stakeholders using a multi-method approach. An awareness survey (n = 87) examined the usage of wearables and general PHI engagement, followed by semi…

AMBEDKAR-A Multi-level Bias Elimination through a Decoding Approach with Knowledge Augmentation for Robust Constitutional Alignment of Language Models
Snehasis Mukhopadhyay, Aryan Kasat, Shivam Dubey, Rahul Karthikeyan, Dhruv Sood, Vinija Jain, Aman Chadha, Amitava Das
https://arxiv.org/abs/2509.02133

AMBEDKAR-A Multi-level Bias Elimination through a Decoding Approach with Knowledge Augmentation for Robust Constitutional Alignment of Language Models
Large Language Models (LLMs) can inadvertently reflect societal biases present in their training data, leading to harmful or prejudiced outputs. In the Indian context, our empirical evaluations across a suite of models reveal that biases around caste and religion are particularly salient. Yet, most existing mitigation strategies are Western-centric and fail to address these local nuances. We propose AMBEDKAR, a framework inspired by the egalitarian vision of Dr B. R. Ambedkar, architect of the …

VULSOVER: Vulnerability Detection via LLM-Driven Constraint Solving
Xiang Li, Yueci Su, Jiahao Liu, Zhiwei Lin, Yuebing Hou, Peiming Gao, Yuanchao Zhang
https://arxiv.org/abs/2509.00882

VULSOVER: Vulnerability Detection via LLM-Driven Constraint Solving
Traditional vulnerability detection methods rely heavily on predefined rule matching, which often fails to capture vulnerabilities accurately. With the rise of large language models (LLMs), leveraging their ability to understand code semantics has emerged as a promising direction for achieving more accurate and efficient vulnerability detection. However, current LLM-based approaches face significant challenges: instability in model outputs, limitations in context length, and hallucination. As a…

ReCode: Improving LLM-based Code Repair with Fine-Grained Retrieval-Augmented Generation
Yicong Zhao, Shisong Chen, Jiacheng Zhang, Zhixu Li
https://arxiv.org/abs/2509.02330 htt…

ReCode: Improving LLM-based Code Repair with Fine-Grained Retrieval-Augmented Generation
Recent advances in large language models (LLMs) have demonstrated impressive capabilities in code-related tasks, such as code generation and automated program repair. Despite their promising performance, most existing approaches for code repair suffer from high training costs or computationally expensive inference. Retrieval-augmented generation (RAG), with its efficient in-context learning paradigm, offers a more scalable alternative. However, conventional retrieval strategies, which are often…

NyayaRAG: Realistic Legal Judgment Prediction with RAG under the Indian Common Law System
Shubham Kumar Nigam, Balaramamahanthi Deepak Patnaik, Shivam Mishra, Ajay Varghese Thomas, Noel Shallum, Kripabandhu Ghosh, Arnab Bhattacharya
https://arxiv.org/abs/2508.00709

NyayaRAG: Realistic Legal Judgment Prediction with RAG under the Indian Common Law System
Legal Judgment Prediction (LJP) has emerged as a key area in AI for law, aiming to automate judicial outcome forecasting and enhance interpretability in legal reasoning. While previous approaches in the Indian context have relied on internal case content such as facts, issues, and reasoning, they often overlook a core element of common law systems, which is reliance on statutory provisions and judicial precedents. In this work, we propose NyayaRAG, a Retrieval-Augmented Generation (RAG) framewo…

Bridging UI Design and chatbot Interactions: Applying Form-Based Principles to Conversational Agents
Sanjay Krishna Anbalagan, Xinrui Nie, Umesh Mohan, Vijay Kumar Kanamarlapudi, Anughna Kommalapati, Xiaodan Zhao
https://arxiv.org/abs/2507.01862

Bridging UI Design and chatbot Interactions: Applying Form-Based Principles to Conversational Agents
Domain specific chatbot applications often involve multi step interactions, such as refining search filters, selecting multiple items, or performing comparisons. Traditional graphical user interfaces (GUIs) handle these workflows by providing explicit "Submit" (commit data) and "Reset" (discard data) actions, allowing back-end systems to track user intent unambiguously. In contrast, conversational agents rely on subtle language cues, which can lead to confusion and incomplete context management…

CyberRAG: An agentic RAG cyber attack classification and reporting tool
Francesco Blefari, Cristian Cosentino, Francesco Aurelio Pironti, Angelo Furfaro, Fabrizio Marozzo
https://arxiv.org/abs/2507.02424

CyberRAG: An agentic RAG cyber attack classification and reporting tool
Intrusion Detection and Prevention Systems (IDS/IPS) in large enterprises can generate hundreds of thousands of alerts per hour, overwhelming security analysts with logs that demand deep, rapidly evolving domain expertise. Conventional machine-learning detectors trim the alert volume but still yield high false-positive rates, while standard single-pass Retrieval-Augmented Generation (RAG) pipelines often retrieve irrelevant context and fail to justify their predictions. To overcome these shortc…

Testing the Untestable? An Empirical Study on the Testing Process of LLM-Powered Software Systems
Cleyton Magalhaes, Italo Santos, Brody Stuart-Verner, Ronnie de Souza Santos
https://arxiv.org/abs/2508.00198

Testing the Untestable? An Empirical Study on the Testing Process of LLM-Powered Software Systems
Background: Software systems powered by large language models are becoming a routine part of everyday technologies, supporting applications across a wide range of domains. In software engineering, many studies have focused on how LLMs support tasks such as code generation, debugging, and documentation. However, there has been limited focus on how full systems that integrate LLMs are tested during development. Aims: This study explores how LLM-powered systems are tested in the context of real-wo…

AI4Research: A Survey of Artificial Intelligence for Scientific Research
Qiguang Chen, Mingda Yang, Libo Qin, Jinhao Liu, Zheng Yan, Jiannan Guan, Dengyun Peng, Yiyan Ji, Hanjing Li, Mengkang Hu, Yimeng Zhang, Yihao Liang, Yuhang Zhou, Jiaqi Wang, Zhi Chen, Wanxiang Che
https://arxiv.org/abs/2507.01903

AI4Research: A Survey of Artificial Intelligence for Scientific Research
Recent advancements in artificial intelligence (AI), particularly in large language models (LLMs) such as OpenAI-o1 and DeepSeek-R1, have demonstrated remarkable capabilities in complex domains such as logical reasoning and experimental coding. Motivated by these advancements, numerous studies have explored the application of AI in the innovation process, particularly in the context of scientific research. These AI technologies primarily aim to develop systems that can autonomously conduct rese…

MOTIF: Modular Thinking via Reinforcement Fine-tuning in LLMs
Purbesh Mitra, Sennur Ulukus
https://arxiv.org/abs/2507.02851 https://a…

MOTIF: Modular Thinking via Reinforcement Fine-tuning in LLMs
Recent advancements in the reasoning capabilities of large language models (LLMs) show that employing group relative policy optimization (GRPO) algorithm for reinforcement learning (RL) training allows the models to use more thinking/reasoning tokens for generating better responses. However, LLMs can generate only a finite amount of tokens while maintaining attention to the previously generated tokens. This limit, also known as the context size of an LLM, is a bottleneck in LLM reasoning with a…

LLMs for Legal Subsumption in German Employment Contracts
Oliver Wardas, Florian Matthes
https://arxiv.org/abs/2507.01734 https://arx…

LLMs for Legal Subsumption in German Employment Contracts
Legal work, characterized by its text-heavy and resource-intensive nature, presents unique challenges and opportunities for NLP research. While data-driven approaches have advanced the field, their lack of interpretability and trustworthiness limits their applicability in dynamic legal environments. To address these issues, we collaborated with legal experts to extend an existing dataset and explored the use of Large Language Models (LLMs) and in-context learning to evaluate the legality of cla…

MMBERT: Scaled Mixture-of-Experts Multimodal BERT for Robust Chinese Hate Speech Detection under Cloaking Perturbations
Qiyao Xue, Yuchen Dou, Ryan Shi, Xiang Lorraine Li, Wei Gao
https://arxiv.org/abs/2508.00760

MMBERT: Scaled Mixture-of-Experts Multimodal BERT for Robust Chinese Hate Speech Detection under Cloaking Perturbations
Hate speech detection on Chinese social networks presents distinct challenges, particularly due to the widespread use of cloaking techniques designed to evade conventional text-based detection systems. Although large language models (LLMs) have recently improved hate speech detection capabilities, the majority of existing work has concentrated on English datasets, with limited attention given to multimodal strategies in the Chinese context. In this study, we propose MMBERT, a novel BERT-based m…

Eka-Eval : A Comprehensive Evaluation Framework for Large Language Models in Indian Languages
Samridhi Raj Sinha, Rajvee Sheth, Abhishek Upperwal, Mayank Singh
https://arxiv.org/abs/2507.01853

Eka-Eval : A Comprehensive Evaluation Framework for Large Language Models in Indian Languages
The rapid advancement of Large Language Models (LLMs) has intensified the need for evaluation frameworks that go beyond English centric benchmarks and address the requirements of linguistically diverse regions such as India. We present EKA-EVAL, a unified and production-ready evaluation framework that integrates over 35 benchmarks, including 10 Indic-specific datasets, spanning categories like reasoning, mathematics, tool use, long-context understanding, and reading comprehension. Compared to e…

Where to show Demos in Your Prompt: A Positional Bias of In-Context Learning
Kwesi Cobbina, Tianyi Zhou
https://arxiv.org/abs/2507.22887 https://arxiv.org/…

Where to show Demos in Your Prompt: A Positional Bias of In-Context Learning
In-context learning (ICL) is a critical emerging capability of large language models (LLMs), enabling few-shot learning during inference by including a few demonstrations (demos) in the prompt. However, it has been found that ICL's performance can be sensitive to the choices of demos and their order. This paper investigates an unexplored new positional bias of ICL for the first time: we observe that the predictions and accuracy can drift drastically when the positions of demos, the system promp…