Tootfinder

Multimodal Language Models with Modality-Specific Experts for Financial Forecasting from Interleaved Sequences of Text and Time Series
Ross Koval, Nicholas Andrews, Xifeng Yan
https://arxiv.org/abs/2509.19628

Multimodal Language Models with Modality-Specific Experts for Financial Forecasting from Interleaved Sequences of Text and Time Series
Text and time series data offer complementary views of financial markets: news articles provide narrative context about company events, while stock prices reflect how markets react to those events. However, despite their complementary nature, effectively integrating these interleaved modalities for improved forecasting remains challenging. In this work, we propose a unified neural architecture that models these interleaved sequences using modality-specific experts, allowing the model to learn u…

I took a look at the changes coming with Python 3.15 – and I can't wait to put them to productive use. I've already updated our tutorials:
• utf-8 as the default encoding: https://python-basics-tutorial.readthedocs.io/en/latest/ty…

Unicode and character encodings
Special characters and escape sequences:\n stands for the newline character and\t for the tab character. Character sequences that begin with a backslash and are used to represent other characters a...

A century of glaciers melting, condensed into a few seconds. Impressive video: https://www.instagram.com/reel/DQWfDRejcPw/?igsh=Zm51d2Qzb2xtNHM4

Michel Galati on Instagram: "A century of change condensed into a few seconds. In collaboration with @glacionaut I transformed historical and contemporary photographs of Swiss and Austrian glaciers into short AI-animated sequences (Dream Machine by @lumalabsai ). Each clip shows the passage from past to present. The climate has always changed, true, but never this fast. In the past two years alone, Swiss glaciers have lost more than 10% of their total volume (source: Swiss Academy of Sciences). What once took centuries now happens in decades. And yes, the music is The Four Seasons by Vivaldi. An homage to nature, and perhaps a reminder of what we stand to lose. 📍 Photos: @glacionaut 🎥 AI Animation: @flashologo 🗓️ Period: ca. 1875–2024 • 🇮🇹 Un secolo di cambiamento racchiuso in pochi secondi. In collaborazione con @glacionaut, ho trasformato fotografie storiche e contemporanee di alcuni ghiacciai svizzeri e austriaci in brevi sequenze animate con l’intelligenza artificiale (Dream Machine di @lumalabsai). Ogni clip mostra il passaggio dal passato al presente. Il clima è sempre cambiato, è vero, ma mai così rapidamente. Negli ultimi due anni i ghiacciai svizzeri hanno perso oltre il 10% del loro volume (fonte: Swiss Academy of Sciences). Ciò che un tempo accadeva in secoli, oggi avviene in decenni. E sì, la musica è Le Quattro Stagioni di Vivaldi. Un omaggio alla natura e, forse, un promemoria di ciò che rischiamo di perdere. 📍 Foto: @glacionaut 🎥 Animazioni AI: @flashologo 🗓️ Periodo: ca. 1875–2024 • • #ai #dreammachine #ray3 #imagetovideo #glaciers #switzerland #austria #climatchange #history"
44K likes, 814 comments - flashologo on October 28, 2025: "A century of change condensed into a few seconds. In collaboration with @glacionaut I transformed historical and contemporary photographs of Swiss and Austrian glaciers into short AI-animated sequences (Dream Machine by @lumalabsai ). Each clip shows the passage from past to present. The climate has always changed, true, but never this fast. In the past two years alone, Swiss glaciers have lost more than 10% of their total volume (sou…

🪐 https://news.harvard.edu/gazette/story/2025/10/you-see-saturns-rings-she-sees-hidden-number-theory/

Some Properties of Homology and Exactness of Nomura's Homology Sequences in a Grandis Homological Category
Yaroslav Kopylov, Vadim Leshkov
https://arxiv.org/abs/2510.11672 h…

Some Properties of Homology and Exactness of Nomura's Homology Sequences in a Grandis Homological Category
We consider Lambek's invariants $\text{Ker}$ and $\text{Im}$ for commutative squares in Grandis homological categories. We prove that Nomura's null sequences exist in such categories and find sufficienct conditions for their exactness. We also prove the coincidence of left and right homology in Grandis homological categories.

Furstenberg systems of certain sequences of superpolynomial growth
Andreu Ferr\'e Moragues, Andreas Koutsogiannis
https://arxiv.org/abs/2510.11957 https://

Furstenberg systems of certain sequences of superpolynomial growth
We give examples of sequences defined by smooth functions of intermediate growth, and we study the Furstenberg systems that model their statistical behavior. In particular, we show that the systems are Bernoulli. We do so by studying exponential sums that reflect the strong equidistribution properties of said sequences. As a by-product of our approach, we also get some convergence results.

These weasels have locally adapted their winter coats to varying snow cover thanks to a million-year-old mutation
#naturalist

Dang it, PostgreSQL is so good.
I did a data export thing. The options I needed were there. Then I needed more options, and then they were also there. They worked the way I guessed. It was fast. It was reliable. Then I thought “oh, what about sequences?!?”, but they already had that covered. My seemingly easy task turned out to be…easy.

Equidistribution of sequences associated with a-points of the derivatives of the Riemann zeta function
Hideki Murahara, Tomokazu Onozuka
https://arxiv.org/abs/2510.07710 https:/…

Equidistribution of sequences associated with a-points of the derivatives of the Riemann zeta function
Fujii investigated the uniform distribution of various sequences associated with the non-trivial zeros of the Riemann zeta function by evaluating certain exponential sums over these zeros. In this paper, we present analogous results for a broader class of functions, establishing the uniform distribution of sequences arising from the $a$-points of the derivatives of the Riemann zeta function.

On Foundation Models for Temporal Point Processes to Accelerate Scientific Discovery
David Berghaus, Patrick Seifner, Kostadin Cvejoski, Ramses J. Sanchez
https://arxiv.org/abs/2510.12640

On Foundation Models for Temporal Point Processes to Accelerate Scientific Discovery
Many scientific fields, from medicine to seismology, rely on analyzing sequences of events over time to understand complex systems. Traditionally, machine learning models must be built and trained from scratch for each new dataset, which is a slow and costly process. We introduce a new approach: a single, powerful model that learns the underlying patterns of event data in context. We trained this "foundation model" on millions of simulated event sequences, teaching it a general-purpose understa…

Compression for Coinductive Infinitary Rewriting: A Generic Approach, with Applications to Cut-Elimination for Non-Wellfounded Proofs
R\'emy Cerda, Alexis Saurin
https://arxiv.org/abs/2510.08420

Compression for Coinductive Infinitary Rewriting: A Generic Approach, with Applications to Cut-Elimination for Non-Wellfounded Proofs
Infinitary rewriting, i.e. rewriting featuring possibly infinite terms and sequences of reduction, is a convenient framework for describing the dynamics of non-terminating but productive rewriting systems. In its original definition based on metric convergence of ordinal-indexed sequences of rewriting steps, a highly desirable property of an infinitary rewriting system is Compression, i.e. the fact that rewriting sequences of arbitrary ordinal length can always be 'compressed' to equivalent seq…

@…
EDIT: nevermind. I converted the references but I'm still open to feedback if you have any.
do you have any advice for me?
I'm surely doing something wrong.
https://

Badinter sollte Vorbild für alle Juristen sein. Er wurde zu Recht im Panthéon aufgenommen.
https://www.franceinfo.fr/politique/mort-de-robert-bad…

VIDEOS. Chanson de Julien Clerc, discours d'Emmanuel Macron... Cinq séquences fortes de la panthéonisation de
L'ancien garde des Sceaux est entré au Panthéon au terme d'une cérémonie qui a attiré un large public le long de la rue Soufflot.

Estimating Sequences with Memory for Minimizing Convex Non-smooth Composite Functions
Endrit Dosti, Sergiy A. Vorobyov, Themistoklis Charalambous
https://arxiv.org/abs/2510.02965

Estimating Sequences with Memory for Minimizing Convex Non-smooth Composite Functions
First-order optimization methods are crucial for solving large-scale data processing problems, particularly those involving convex non-smooth composite objectives. For such problems with convex non-smooth composite objectives, we introduce a new class of generalized composite estimating sequences, devised by exploiting the information embedded in the iterates generated during the minimization process. Building on these sequences, we propose a novel accelerated first-order method tailored for su…

Linear Recurrent Sequences, Markov Chains and Their Applications in Graph Theory
Rebecca Carter, M. Ram Murty
https://arxiv.org/abs/2510.03537 https://arxi…

Linear Recurrent Sequences, Markov Chains and Their Applications in Graph Theory
After a brief review of the key theorems concerning recurrent sequences, we give an explicit computation of the inverse of the Vandermonde matrix. This will then be used to derive sub-exponential decay error terms in the ergodic theorem of Markov chains. Finally, we apply these results to give estimates for the diameters of directed graphs.

Emergent Mixed States for Baby Universes and Black Holes
Jonah Kudler-Flam, Edward Witten
https://arxiv.org/abs/2510.06376 https://arxiv.org/pdf/2510.06376…

Emergent Mixed States for Baby Universes and Black Holes
We examine the behavior of sequences of states in the large $N$ limit of AdS/CFT duality in cases in which the bulk duals involve baby universes or black holes. Such sequences generally fail to converge as pure states. Under suitable conditions, such as diverging coarse-grained entropy, they can converge to mixed states for the large $N$ algebra, as in the case of black holes. For Euclidean preparations that produce baby universes, the sequences do not converge, due to wormhole contributions, a…

Multi-Granularity Sequence Denoising with Weakly Supervised Signal for Sequential Recommendation
Liang Li, Zhou Yang, Xiaofei Zhu
https://arxiv.org/abs/2510.10564 https://

Multi-Granularity Sequence Denoising with Weakly Supervised Signal for Sequential Recommendation
Sequential recommendation aims to predict the next item based on user interests in historical interaction sequences. Historical interaction sequences often contain irrelevant noisy items, which significantly hinders the performance of recommendation systems. Existing research employs unsupervised methods that indirectly identify item-granularity irrelevant noise by predicting the ground truth item. Since these methods lack explicit noise labels, they are prone to misidentify users' interested i…

Retrieving the structure of probabilistic sequences from EEG data during the goalkeeper game
P. R. Cabral-Passos, P. S. Azevedo, V. H. Moraes, B. L. Ramalho, A. Duarte, C. D. Vargas
https://arxiv.org/abs/2510.06344

Retrieving the structure of probabilistic sequences from EEG data during the goalkeeper game
This work draws on the conjecture that fingerprints of stochastic event sequences can be retrieved from electroencephalographic data (EEG) recorded during a behavioral task. To test this, we used the Goalkeeper Game (game.numec.prp.usp.br). Acting as a goalkeeper, the participant predicted each kick in a probabilistic sequence while EEG activity was recorded. At each trial, driven by a context tree, the kicker chose one of three options: left, center, or right. The goalkeeper then predicted the…

Spectral theory of matrix-sequences: perspectives of the GLT analysis and beyond
Stefano Serra-Capizzano
https://arxiv.org/abs/2509.19544 https://arxiv.org…

Spectral theory of matrix-sequences: perspectives of the GLT analysis and beyond
In recent years there has been a growing attention on distribution results in the sense of Weyl for the collective behavior of eigenvalues and singular values of matrix-sequences. Starting from the work of Szegö regarding the case of Toeplitz matrix-sequences, there has been a wealth of associated results, which culminated in the works of Tilli and of Tyrtyshnikov, Zamarashkin, and of the author for preconditioned and non-preconditioned $r$-block $d$-level Toeplitz matrix-sequences with Lebesg…

Metric Convergence of Sequences of Static Spacetimes with the Null Distance
Brian Allen
https://arxiv.org/abs/2510.02237 https://arxiv.org/pdf/2510.02237…

Metric Convergence of Sequences of Static Spacetimes with the Null Distance
How should one define metric space notions of convergence for sequences of spacetimes? Since a Lorentzian manifold does not define a metric space directly, the uniform convergence, Gromov-Hausdorff (GH) convergence, and Sormani-Wenger Intrinsic Flat (SWIF) convergence does not extend automatically. One approach is to define a metric space structure, which is compatible with the Lorentzian structure, so that the usual notions of convergence apply. This approach was taken by C. Sormani and C. Veg…

Autoencoding Coordinate Sequences from Psychophysiologic Signals
Timothy L. Hutcheson, Anil K. Raj
https://arxiv.org/abs/2510.07415 https://arxiv.org/pdf/2…

Autoencoding Coordinate Sequences from Psychophysiologic Signals
We present a method for converting 24 channels of psychophysiologic time series data collected from individual participants via electroencephalogram (EEG), electrocardiogram (ECG), electrodermal activity (EDA), respiration rate (RR) into trackable three dimensional (3D) coordinates sufficient to estimate participation in specific task and cognitive states.

"Scientists develop new floating box array for offshore PV"
#SolarPower #Energy #Renewables

Scientists develop new floating box array for offshore PV
Researchers have tested a 9 x 11 floating box array for offshore photovoltaics in a large wave flume, under regular and irregular wave sequences. They have placed wireless gyroscopes on three of the boxes, representing the front, middle, and rear positions.

Replaced article(s) found for cs.AI. https://arxiv.org/list/cs.AI/new
[12/14]:
- Clip Your Sequences Fairly: Enforcing Length Fairness for Sequence-Level RL
Hanyi Mao, Quanjia Xiao, Lei Pang, Haixiao Liu

On Diophantine equations involving intersection of Thabit and Williams numbers base $b$ and some ternary recurrent sequences
Bibhu Prasad Tripathy, Asutosh Satapathy, Utkal Keshari Dutta, Bijan Kumar Patel
https://arxiv.org/abs/2510.12145

On Diophantine equations involving intersection of Thabit and Williams numbers base $b$ and some ternary recurrent sequences
Let $\mathcal{P}_{n}$ be the $n$-th Padovan number, $E_{n}$ be the $n$-th Perrin number and $N_{n}$ be the $n$-th Narayana's cows number. Let $b$ be a positive integer such that $b \geq 2$. In this paper, we study the Diophantine equations \[ \mathcal{P}_{n} = (b \pm 1)\cdot b^{l} \pm 1, \] \[ E_{n} = (b \pm 1)\cdot b^{l} \pm 1, \] and \[ N_{n} = (b \pm 1)\cdot b^{l} \pm 1, \] in non-negative integers $n, b$ and positive integer $l$. As a result, we determine the Padovan, Perrin…

GramTrans: A Better Code Representation Approach in Code Generation
Zhao Zhang, Qingyuan Liang, Zeyu Sun, Yizhou Chen, Guoqing Wang, Yican Sun, Lu Zhang, Ge Li, Yingfei Xiong
https://arxiv.org/abs/2510.02887

GramTrans: A Better Code Representation Approach in Code Generation
Code generation has shown great promise in assisting software development. A fundamental yet underexplored question is how the choice of code representation affects model performance. While existing studies employ various representations, such as treating code as plain text, grammar rule sequences, or syntax tree sequences, they lack a principled understanding of the relationship between parsing difficulty and model effectiveness. This paper proposes a conjecture: the easier a representation is…

InfVSR: Breaking Length Limits of Generic Video Super-Resolution
Ziqing Zhang, Kai Liu, Zheng Chen, Xi Li, Yucong Chen, Bingnan Duan, Linghe Kong, Yulun Zhang
https://arxiv.org/abs/2510.00948

InfVSR: Breaking Length Limits of Generic Video Super-Resolution
Real-world videos often extend over thousands of frames. Existing video super-resolution (VSR) approaches, however, face two persistent challenges when processing long sequences: (1) inefficiency due to the heavy cost of multi-step denoising for full-length sequences; and (2) poor scalability hindered by temporal decomposition that causes artifacts and discontinuities. To break these limits, we propose InfVSR, which novelly reformulates VSR as an autoregressive-one-step-diffusion paradigm. This…

STaTS: Structure-Aware Temporal Sequence Summarization via Statistical Window Merging
Disharee Bhowmick, Ranjith Ramanathan, Sathyanarayanan N. Aakur
https://arxiv.org/abs/2510.09593

STaTS: Structure-Aware Temporal Sequence Summarization via Statistical Window Merging
Time series data often contain latent temporal structure, transitions between locally stationary regimes, repeated motifs, and bursts of variability, that are rarely leveraged in standard representation learning pipelines. Existing models typically operate on raw or fixed-window sequences, treating all time steps as equally informative, which leads to inefficiencies, poor robustness, and limited scalability in long or noisy sequences. We propose STaTS, a lightweight, unsupervised framework for …

Psyzkaller: Learning from Historical and On-the-Fly Execution Data for Smarter Seed Generation in OS kernel Fuzzing
Boyu Liu, Yang Zhang, Liang Cheng, Yi Zhang, Junjie Fan, Yu Fu
https://arxiv.org/abs/2510.08918

Psyzkaller: Learning from Historical and On-the-Fly Execution Data for Smarter Seed Generation in OS kernel Fuzzing
Fuzzing has become a cornerstone technique for uncovering vulnerabilities and enhancing the security of OS kernels. However, state-of-the-art kernel fuzzers, including the de facto standard Syzkaller, struggle to generate valid syscall sequences that respect implicit Syscall Dependency Relations (SDRs). Consequently, many generated seeds either fail kernel validation or cannot penetrate deep execution paths, resulting in significant inefficiency. We hypothesize that SDRs can be effectively le…

Raiders report: Pete Carroll defiant when asked if team has quit on him https://www.reviewjournal.com/sports/raiders/raiders-report-pete-carroll-defiant-when-asked-if-team-has-quit-on-him-3595605/

Learning Hanzi Character Through VR-Based Mortise-Tenon
Conglin Ma, Jiatong Li, Sen-Zhe Xu, Ju Dai, Jie Liu, Feng Zhou
https://arxiv.org/abs/2510.11264 https://

Learning Hanzi Character Through VR-Based Mortise-Tenon
This paper introduces a novel VR-based system that redefines the acquisition of Hanzi character literacy by integrating traditional mortise-tenon joinery principles (HVRMT).Addressing the challenge of abstract character memorization in digital learning,our system deconstructs Hanzi components into interactive "structural radicals"akin to wooden joint modules.Leveraging PICO's 6DoF spatial tracking and LLM's morphological analysis,learners assemble stroke sequences with haptic feedback simulatin…

Adaptive Science Operations in Deep Space Missions Using Offline Belief State Planning
Grace Ra Kim, Hailey Warner, Duncan Eddy, Evan Astle, Zachary Booth, Edward Balaban, Mykel J. Kochenderfer
https://arxiv.org/abs/2510.08812

Adaptive Science Operations in Deep Space Missions Using Offline Belief State Planning
Deep space missions face extreme communication delays and environmental uncertainty that prevent real-time ground operations. To support autonomous science operations in communication-constrained environments, we present a partially observable Markov decision process (POMDP) framework that adaptively sequences spacecraft science instruments. We integrate a Bayesian network into the POMDP observation space to manage the high-dimensional and uncertain measurements typical of astrobiology missions…

Further analysis of Peeling Sequences
D\'aniel G\'abor Simon
https://arxiv.org/abs/2510.03832 https://arxiv.org/pdf/2510.03832

Further analysis of Peeling Sequences
Let $P\subset \mathbf{R}^2$ be a set of $n$ points in general position. A peeling sequence of $P$ is a list of its points, such that if we remove the points from $P$ in that order, we always remove the next point from the convex hull of the remainder of $P$. Using the methodology of Dumitrescu and Tóth \cite{Dumitrescu}, with more careful analysis, we improve the upper bound on the minimum number of peeling sequences for an $n$ point set in the plane from $\frac{12.29^n}{100}$ to $\frac{9.78^n…

PRISM: Enhancing Protein Inverse Folding through Fine-Grained Retrieval on Structure-Sequence Multimodal Representations
Sazan Mahbub, Souvik Kundu, Eric P. Xing
https://arxiv.org/abs/2510.11750

PRISM: Enhancing Protein Inverse Folding through Fine-Grained Retrieval on Structure-Sequence Multimodal Representations
Designing protein sequences that fold into a target three-dimensional structure, known as the inverse folding problem, is central to protein engineering but remains challenging due to the vast sequence space and the importance of local structural constraints. Existing deep learning approaches achieve strong recovery rates, yet they lack explicit mechanisms to reuse fine-grained structure-sequence patterns that are conserved across natural proteins. We present PRISM, a multimodal retrieval-augme…

Just pushed an update for https://thi.ng/genart-api (v0.29.0) which now supports dynamic switching of time providers. This is useful for situations where you want to switch from realtime animation to offline-based timing, e.g. to export high-resolution image sequences and give the browser time to grab & encode e…

thi.ng/genart-api

Signed null sequences and Hausdorff dimension
Rich\'ard Balka, Korn\'elia H\'era, Gergely Kiss
https://arxiv.org/abs/2509.20181 https://arxiv.o…

Signed null sequences and Hausdorff dimension
We investigate the convergence of signed null sequences of the form \[ \sum_{n=1}^\infty \varepsilon_n a_n, \quad \varepsilon_n \in \{-1,1\}, \] where $(a_n)$ tends to zero in $\mathbb{R}^d$. Our main result shows that for any such sequence, the set of sign sequences yielding convergence has full Hausdorff dimension in the natural ultrametric topology. This answers a question of Mattila in the one-dimensional case, for which we provide an elementary proof. Moreover, if $(a_n)\notin \ell^1$ in o…

On minimal product-one sequences of maximal length over the non-abelian group of order $pq$
Danilo Vilela Avelar, Fabio Enrique Brochero Mart\'inez, S\'avio Ribas
https://arxiv.org/abs/2510.00070

On minimal product-one sequences of maximal length over the non-abelian group of order $pq$
Let $G$ be a finite group. A sequence over $G$ is a finite multiset of elements of $G$, and it is called product-one if its terms can be ordered so that their product is the identity of $G$. The large Davenport constant $\D(G)$ is the maximal length of a minimal product-one sequence, that is, a product-one sequence that cannot be partitioned into two nontrivial product-one subsequences. Let $p,q$ be odd prime numbers with $p \mid q-1$ and let $C_q \rtimes C_p$ denote the non-abelian group of or…

Event Horizon Telescope Pattern Speeds in the Visibility Domain
Nicholas S. Conroy, Michi Baub\"ock, Vedant Dhruv, Daeyoung Lee, Chi-kwan Chan, Abhishek V. Joshi, Ben Prather, Charles F. Gammie
https://arxiv.org/abs/2510.08848

Event Horizon Telescope Pattern Speeds in the Visibility Domain
The Event Horizon Telescope is preparing to produce time sequences of black hole images, or movies. In anticipation, we developed an autocorrelation technique to measure apparent rotational motion using the image-domain pattern speed $Ω_p$. Here, we extend this technique to the visibility domain and introduce the visibility amplitude pattern speed $Ω_{\mathrm{VA}}$. We show that in the Illinois v3 library of EHT source models, $Ω_{\mathrm{VA}}$ depends on the source inclination, black hole m…

Targeted Sequential Pattern Mining with High Average Utility
Kai Cao, Yucong Duan, Wensheng Gan
https://arxiv.org/abs/2510.10115 https://arxiv.org/pdf/2510…

Targeted Sequential Pattern Mining with High Average Utility
Incorporating utility into targeted pattern mining can address the practical limitations of traditional frequency-based approaches. However, utility-based methods often suffer from generating a large number of long and complicated sequences. To improve pattern relevance and interpretability, average utility provides a more balanced metric by considering both utility and sequence length. Moreover, incorporating user-defined query targets into the mining process enhances usability and interactivi…

Denominators of R-matrices, higher Dorey's rules and a generalization of T-systems for quantum affine algebras
Se-jin Oh, Travis Scrimshaw
https://arxiv.org/abs/2510.10874 h…

Denominators of R-matrices, higher Dorey's rules and a generalization of T-systems for quantum affine algebras
We construct a higher level analogue of Dorey's rule, which describe certain surjective morphisms between Kirillov--Reshetikhin (KR) modules over quantum affine algebras. Building on this, we establish a generalized T-system of short exact sequences and prove the denominator formula between KR modules in all nonexceptional types, except with only mild ambiguities persisting in type $C_n^{(1)}$. As a consequence, we can completely classify when a tensor product of KR modules is simple. These res…

BurstEngine: an Efficient Distributed Framework for Training Transformers on Extremely Long Sequences of over 1M Tokens
Ao Sun, Weilin Zhao, Xu Han, Cheng Yang, Zhiyuan Liu, Chuan Shi, Maosong sun
https://arxiv.org/abs/2509.19836

BurstEngine: an Efficient Distributed Framework for Training Transformers on Extremely Long Sequences of over 1M Tokens
Existing methods for training LLMs on long-sequence data, such as Tensor Parallelism and Context Parallelism, exhibit low Model FLOPs Utilization as sequence lengths and number of GPUs increase, especially when sequence lengths exceed 1M tokens. To address these challenges, we propose BurstEngine, an efficient framework designed to train LLMs on long-sequence data. BurstEngine introduces BurstAttention, an optimized distributed attention with lower communication cost than RingAttention. BurstAt…

Replaced article(s) found for cs.MS. https://arxiv.org/list/cs.MS/new
[1/1]:
- QMCPy: A Python Software for Randomized Low-Discrepancy Sequences, Quasi-Monte Carlo, and Fast Ke...
Aleksei G Sorokin

A Class of Functionals on the Sequence Space $s$ Satisfying the Palais-Smale Condition
Kaveh Eftekharinasab
https://arxiv.org/abs/2510.10146 https://arxiv.…

A Class of Functionals on the Sequence Space $s$ Satisfying the Palais-Smale Condition
We introduce a class of functionals on the space of rapidly decreasing sequences $s$, called $\mathcal{F}_s$-functionals, defined as decomposable sums of quadratic and convex terms with quadratic growth. We prove that such functionals satisfy the Palais-Smale condition and admit a unique global minimum. Furthermore, we show that the Palais-Smale condition is preserved under linear homeomorphisms. This allows us to construct corresponding functionals satisfying the Palais-Smale condition on Fré…

To do: the representative image of an image sequence is the image where it’s pixel vector has the biggest cosine similarity to the average pixel vector of all images of the sequences. O(n)

The condensed homotopy type of a scheme
Peter J. Haine, Tim Holzschuh, Marcin Lara, Catrin Mair, Louis Martini, Sebastian Wolf with an appendix by Bogdan Zavyalov
https://arxiv.org/abs/2510.07443

The condensed homotopy type of a scheme
We study a condensed version of the étale homotopy type of a scheme, which refines both the usual étale homotopy type of Friedlander-Artin-Mazur and the proétale fundamental group of Bhatt-Scholze. In the first part of this paper, we prove that this condensed homotopy type satisfies descent along integral morphisms and that the expected fiber sequences hold. We also provide explicit computations, for example, for rings of continuous functions. A key ingredient in many of our arguments is a d…

Cache-to-Cache: Direct Semantic Communication Between Large Language Models
Tianyu Fu, Zihan Min, Hanling Zhang, Jichao Yan, Guohao Dai, Wanli Ouyang, Yu Wang
https://arxiv.org/abs/2510.03215

Cache-to-Cache: Direct Semantic Communication Between Large Language Models
Multi-LLM systems harness the complementary strengths of diverse Large Language Models, achieving performance and efficiency gains unattainable by a single model. In existing designs, LLMs communicate through text, forcing internal representations to be transformed into output token sequences. This process both loses rich semantic information and incurs token-by-token generation latency. Motivated by these limitations, we ask: Can LLMs communicate beyond text? Oracle experiments show that enric…

GraphTracer: Graph-Guided Failure Tracing in LLM Agents for Robust Multi-Turn Deep Search
Heng Zhang, Yuling Shi, Xiaodong Gu, Haochen You, Zijian Zhang, Lubin Gan, Yilei Yuan, Jin Huang
https://arxiv.org/abs/2510.10581

GraphTracer: Graph-Guided Failure Tracing in LLM Agents for Robust Multi-Turn Deep Search
Multi-agent systems powered by Large Language Models excel at complex tasks through coordinated collaboration, yet they face high failure rates in multi-turn deep search scenarios. Existing temporal attribution methods struggle to accurately diagnose root causes, particularly when errors propagate across multiple agents. Attempts to automate failure attribution by analyzing action sequences remain ineffective due to their inability to account for information dependencies that span agents. This …

Replaced article(s) found for cs.SC. https://arxiv.org/list/cs.SC/new
[1/1]:
- Computing with D-Algebraic Sequences
Bertrand Teguia Tabuguia
https://

Coarse scrambling for Sobol' and Niederreiter sequences
Kosuke Suzuki
https://arxiv.org/abs/2510.02111 https://arxiv.org/pdf/2510.02111

Coarse scrambling for Sobol' and Niederreiter sequences
We introduce \emph{coarse scrambling}, a novel randomization for digital sequences that permutes blocks of digits in a mixed-radix representation. This construction is designed to preserve the powerful $(0,\boldsymbol{e},d)$-sequence property of the underlying points. For sufficiently smooth integrands, we prove that this method achieves the canonical $O(n^{-3+ε})$ variance decay rate, matching that of standard Owen's scrambling. Crucially, we show that its maximal gain coefficient grows only …

Neptune: Advanced ML Operator Fusion for Locality and Parallelism on GPUs
Yifan Zhao, Egan Johnson, Prasanth Chatarasi, Vikram Adve, Sasa Misailovic
https://arxiv.org/abs/2510.08726

Neptune: Advanced ML Operator Fusion for Locality and Parallelism on GPUs
Operator fusion has become a key optimization for deep learning, which combines multiple deep learning operators to improve data reuse and reduce global memory transfers. However, existing tensor compilers struggle to fuse complex reduction computations involving loop-carried dependencies, such as attention mechanisms. The paper introduces Neptune, a tensor compiler for advanced operator fusion for sequences of reduction operators. Neptune presents a new approach for advanced operator fusion,…

Replaced article(s) found for cs.FL. https://arxiv.org/list/cs.FL/new
[1/1]:
- Completing the picture for the Skolem Problem on order-4 linear recurrence sequences
Piotr Bacik

Local degrees of FI-modules through spectral sequences
Cihan Bahran
https://arxiv.org/abs/2509.23998 https://arxiv.org/pdf/2509.23998

Local degrees of FI-modules through spectral sequences
We establish bounds on the local degrees of FI-modules (in the sense of Church-Miller-Nagpal-Reinhold [CMNR18]) through a spectral sequence which only depend on the initial page and is independent from how many pages have been flipped.

HIV-1 protease cleavage sites detection with a Quantum convolutional neural network algorithm
Junggu Choi, Junho Lee, Kyle L. Jung, Jae U. Jung
https://arxiv.org/abs/2510.01993 …

HIV-1 protease cleavage sites detection with a Quantum convolutional neural network algorithm
In this study, we propose a quantum convolutional neural network (QCNN)-based framework with the neural quantum embedding (NQE) to predict HIV-1 protease cleavage sites in amino acid sequences from viral and human proteins. To assess the effectiveness and robustness of our framework, we compared the classification performance against classical neural networks under both noiseless and noisy simulations. Among experimental conditions, the QCNN with the angle and amplitude encoding NQE conditions …

A Hierarchical Quantized Tokenization Framework for Task-Adaptive Graph Representation Learning
Yang Xiang, Li Fan, Chenke Yin, Chengtao Ji
https://arxiv.org/abs/2510.12369 http…

A Hierarchical Quantized Tokenization Framework for Task-Adaptive Graph Representation Learning
Recent progress in language and vision foundation models demonstrates the importance of discrete token interfaces that transform complex inputs into compact sequences for large-scale modeling. Extending this paradigm to graphs requires a tokenization scheme that handles non-Euclidean structures and multi-scale dependencies efficiently. Existing approaches to graph tokenization, linearized, continuous, and quantized, remain limited in adaptability and efficiency. In particular, most current quan…

PSTTS: A Plug-and-Play Token Selector for Efficient Event-based Spatio-temporal Representation Learning
Xiangmo Zhao, Nan Yang, Yang Wang, Zhanwen Liu
https://arxiv.org/abs/2509.22481

PSTTS: A Plug-and-Play Token Selector for Efficient Event-based Spatio-temporal Representation Learning
Mainstream event-based spatio-temporal representation learning methods typically process event streams by converting them into sequences of event frames, achieving remarkable performance. However, they neglect the high spatial sparsity and inter-frame motion redundancy inherent in event frame sequences, leading to significant computational overhead. Existing token sparsification methods for RGB videos rely on unreliable intermediate token representations and neglect the influence of event noise…

Same model, better performance: the impact of shuffling on DNA Language Models benchmarking
Davide Greco, Konrad Rawlik
https://arxiv.org/abs/2510.12617 https://

Same model, better performance: the impact of shuffling on DNA Language Models benchmarking
Large Language Models are increasingly popular in genomics due to their potential to decode complex biological sequences. Hence, researchers require a standardized benchmark to evaluate DNA Language Models (DNA LMs) capabilities. However, evaluating DNA LMs is a complex task that intersects genomic's domain-specific challenges and machine learning methodologies, where seemingly minor implementation details can significantly compromise benchmark validity. We demonstrate this through BEND (Benchm…

Linear RNNs for autoregressive generation of long music samples
Konrad Szewczyk, Daniel Gallo Fern\'andez, James Townsend
https://arxiv.org/abs/2510.02401 https://

Linear RNNs for autoregressive generation of long music samples
Directly learning to generate audio waveforms in an autoregressive manner is a challenging task, due to the length of the raw sequences and the existence of important structure on many different timescales. Traditional approaches based on recurrent neural networks, as well as causal convolutions and self-attention, have only had limited success on this task. However, recent work has shown that deep state space models, also referred to as linear RNNs, can be highly efficient in this context. In …

The Lie of the Average: How Class Incremental Learning Evaluation Deceives You?
Guannan Lai, Da-Wei Zhou, Xin Yang, Han-Jia Ye
https://arxiv.org/abs/2509.22580 https://

The Lie of the Average: How Class Incremental Learning Evaluation Deceives You?
Class Incremental Learning (CIL) requires models to continuously learn new classes without forgetting previously learned ones, while maintaining stable performance across all possible class sequences. In real-world settings, the order in which classes arrive is diverse and unpredictable, and model performance can vary substantially across different sequences. Yet mainstream evaluation protocols calculate mean and variance from only a small set of randomly sampled sequences. Our theoretical anal…

A Digital Pheromone-Based Approach for In/Out-of-Control Classification
Pedro Pestana, M. F\'atima Brilhante
https://arxiv.org/abs/2510.07329 https://a…

A Digital Pheromone-Based Approach for In/Out-of-Control Classification
In complex production lines, it is essential to have strict, fast-acting rules to determine whether the system is In Control (InC) or Out of Control (OutC). This study explores a bio-inspired method that digitally mimics ant colony behavior to classify InC/OutC states and forecast imminent transitions requiring maintenance. A case study on industrial potato chip frying provides the application context. During each two-minute frying cycle, sequences of eight temperature readings are collected. E…

Sequences of odd length in strict partitions IV: the combinatorics of parameterized Rogers-Ramanujan type identities
Haijun Li
https://arxiv.org/abs/2509.24942 https://

Sequences of odd length in strict partitions IV: the combinatorics of parameterized Rogers-Ramanujan type identities
In the first three papers, we conducted a series of discussions on the statistics of strict partitions and Rogers-Ramanujan partitions, specifically the sequences of odd length (denoted as $\mathrm{sol}$) and its extensions. We established bijections for some Rogers-Ramanujan type identities. This paper will continue that series of work, and first we will use the bijective method to re-establish several parameterized Rogers-Ramanujan type identities, which appeared in the recent work of Hao-Kua…

Lacunary sequences whose reciprocal sums represent all rationals in an interval
Wouter van Doorn, Vjekoslav Kova\v{c}
https://arxiv.org/abs/2509.24971 https://

Lacunary sequences whose reciprocal sums represent all rationals in an interval
Answering a question of Michael Bleicher and Paul Erdős, we show that there exists a lacunary sequence of positive integers such that finite sums of reciprocals of its terms attain all rational numbers from a non-empty open interval. We also study several stronger variants of their original problem: determining the value of the optimal lacunarity parameter, representing rationals infinitely many times, finding such lacunary sequences with arbitrarily large jumps, and relating the maximal lengt…

A topology for Engel expansions: evaluation and digit coding maps
Min Woong Ahn
https://arxiv.org/abs/2510.10850 https://arxiv.org/pdf/2510.10850

A topology for Engel expansions: evaluation and digit coding maps
We develop a topological framework for Engel expansions that treats both directions of the correspondence between points of $(0,1]$ and nondecreasing digit sequences. We endow the sequence space with the product topology to study the evaluation map, and we fix a nonterminating digit algorithm to study the digit coding map. We also record the correspondence between cylinder sets and fundamental intervals, and give an application to Baire category results for functions of the digits.

Automaton Constrained Q-Learning
Anastasios Manganaris, Vittorio Giammarino, Ahmed H. Qureshi
https://arxiv.org/abs/2510.05061 https://arxiv.org/pdf/2510.0…

Automaton Constrained Q-Learning
Real-world robotic tasks often require agents to achieve sequences of goals while respecting time-varying safety constraints. However, standard Reinforcement Learning (RL) paradigms are fundamentally limited in these settings. A natural approach to these problems is to combine RL with Linear-time Temporal Logic (LTL), a formal language for specifying complex, temporally extended tasks and safety constraints. Yet, existing RL methods for LTL objectives exhibit poor empirical performance in compl…

Adaptive Protein Design Protocols and Middleware
Aymen Alsaadi, Jonathan Ash, Mikhail Titov, Matteo Turilli, Andre Merzky, Shantenu Jha, Sagar Khare
https://arxiv.org/abs/2510.06396

Adaptive Protein Design Protocols and Middleware
Computational protein design is experiencing a transformation driven by AI/ML. However, the range of potential protein sequences and structures is astronomically vast, even for moderately sized proteins. Hence, achieving convergence between generated and predicted structures demands substantial computational resources for sampling. The Integrated Machine-learning for Protein Structures at Scale (IMPRESS) offers methods and advanced computing systems for coupling AI to high-performance computing…

Z-scores-based methods and their application to biological monitoring: An extended analysis of professional soccer players and cyclists athletes
Geoffroy C. B. Berthelot (IRMES - URP\_7329, RELAIS), Brigitte Gelein (ENSAI, IRMAR), Eric Meinadier (FFC), Emmanuel Orhant (FFF), J\'er\^ome Dedecker (MAP5 - UMR 8145)
https://arxiv.org/abs/2…

Z-scores-based methods and their application to biological monitoring: An extended analysis of professional soccer players and cyclists athletes
The increase in the collection of biological data allows for the individual and longitudinal monitoring of hematological or urine biomarkers. However, identifying abnormal behavior in these biological sequences is not trivial. Moreover, the complexity of the biological data (correlation between biomarkers, seasonal effects, etc.) is also an issue. Z-score methods can help assess the abnormality in these longitudinal sequences while capturing some features of the biological complexity. This work…

Synthetic Dialogue Generation for Interactive Conversational Elicitation & Recommendation (ICER)
Moonkyung Ryu, Chih-Wei Hsu, Yinlam Chow, Mohammad Ghavamzadeh, Craig Boutilier
https://arxiv.org/abs/2510.02331

Synthetic Dialogue Generation for Interactive Conversational Elicitation & Recommendation (ICER)
While language models (LMs) offer great potential for conversational recommender systems (CRSs), the paucity of public CRS data makes fine-tuning LMs for CRSs challenging. In response, LMs as user simulators qua data generators can be used to train LM-based CRSs, but often lack behavioral consistency, generating utterance sequences inconsistent with those of any real user. To address this, we develop a methodology for generating natural dialogues that are consistent with a user's underlying sta…

The classification of quasi-elliptic fibrations and unexpected plane cubics in characteristics 2 and 3
Jake Kettinger
https://arxiv.org/abs/2510.06365 https://

The classification of quasi-elliptic fibrations and unexpected plane cubics in characteristics 2 and 3
In this paper, we categorize all isomorphism classes of quasi-elliptic surfaces over a field $k$ of characteristic 2 or 3. For every quasi-elliptic surface $X$, we classify all possible sequences of blow-downs from $X$ to the projective plane $\mathbb{P}^2_k$. We then use these categorizations to identify all unexpected plane cubic curves in characteristic 2 and present a proof of the lack of unexpected cubics in characteristic 3. Before the work in this paper -- based partly on the author's th…

ARTDECO: Towards Efficient and High-Fidelity On-the-Fly 3D Reconstruction with Structured Scene Representation
Guanghao Li, Kerui Ren, Linning Xu, Zhewen Zheng, Changjian Jiang, Xin Gao, Bo Dai, Jian Pu, Mulin Yu, Jiangmiao Pang
https://arxiv.org/abs/2510.08551

ARTDECO: Towards Efficient and High-Fidelity On-the-Fly 3D Reconstruction with Structured Scene Representation
On-the-fly 3D reconstruction from monocular image sequences is a long-standing challenge in computer vision, critical for applications such as real-to-sim, AR/VR, and robotics. Existing methods face a major tradeoff: per-scene optimization yields high fidelity but is computationally expensive, whereas feed-forward foundation models enable real-time inference but struggle with accuracy and robustness. In this work, we propose ARTDECO, a unified framework that combines the efficiency of feed-forw…

Verification of Sequential Convex Programming for Parametric Non-convex Optimization
Rajiv Sambharya, Nikolai Matni, George Pappas
https://arxiv.org/abs/2511.10622 https://arxiv.org/pdf/2511.10622 https://arxiv.org/html/2511.10622
arXiv:2511.10622v1 Announce Type: new
Abstract: We introduce a verification framework to exactly verify the worst-case performance of sequential convex programming (SCP) algorithms for parametric non-convex optimization. The verification problem is formulated as an optimization problem that maximizes a performance metric (e.g., the suboptimality after a given number of iterations) over parameters constrained to be in a parameter set and iterate sequences consistent with the SCP update rules. Our framework is general, extending the notion of SCP to include both conventional variants such as trust-region, convex-concave, and prox-linear methods, and algorithms that combine convex subproblems with rounding steps, as in relaxing and rounding schemes. Unlike existing analyses that may only provide local guarantees under limited conditions, our framework delivers global worst-case guarantees--quantifying how well an SCP algorithm performs across all problem instances in the specified family. Applications in control, signal processing, and operations research demonstrate that our framework provides, for the first time, global worst-case guarantees for SCP algorithms in the parametric setting.
toXiv_bot_toot

Cross-attention Secretly Performs Orthogonal Alignment in Recommendation Models
Hyunin Lee, Yong Zhang, Hoang Vu Nguyen, Xiaoyi Liu, Namyong Park, Christopher Jung, Rong Jin, Yang Wang, Zhigang Wang, Somayeh Sojoudi, Xue Feng
https://arxiv.org/abs/2510.09435

Cross-attention Secretly Performs Orthogonal Alignment in Recommendation Models
Cross-domain sequential recommendation (CDSR) aims to align heterogeneous user behavior sequences collected from different domains. While cross-attention is widely used to enhance alignment and improve recommendation performance, its underlying mechanism is not fully understood. Most researchers interpret cross-attention as residual alignment, where the output is generated by removing redundant and preserving non-redundant information from the query input by referencing another domain data whic…

Breaking the Likelihood Trap: Consistent Generative Recommendation with Graph-structured Model
Qiya Yang, Xiaoxi Liang, Zeping Xiao, Yingjie Deng, Yalong Wang, Yongqi Liu, Han Li
https://arxiv.org/abs/2510.10127

Breaking the Likelihood Trap: Consistent Generative Recommendation with Graph-structured Model
Reranking, as the final stage of recommender systems, demands real-time inference, accuracy, and diversity. It plays a crucial role in determining the final exposure, directly influencing user experience. Recently, generative reranking has gained increasing attention for its strong ability to model complex dependencies among items. However, most existing methods suffer from the "likelihood trap", where high-likelihood sequences are often perceived as low-quality by humans. These models tend to …

On Finite Pseudorandom Binary Sequences: Generalized polynomials
Manfred G. Madritsch, Robert F. Tichy
https://arxiv.org/abs/2509.19810 https://arxiv.org/p…

On Finite Pseudorandom Binary Sequences: Generalized polynomials
In the present paper we generate binary pseudorandom sequences using generalized polynomials. A generalized polynomial is a function in whose description we not only allow addition and product (as it is the case in usual polynomials) but also the floor function. We estimate the well-distribution measure, looking at the ``randomness'' along arithmetic progressions.

Crosslisted article(s) found for q-bio.NC. https://arxiv.org/list/q-bio.NC/new
[1/1]:
- Autoencoding Coordinate Sequences from Psychophysiologic Signals
Timothy L. Hutcheson, Anil K. Raj

Degree sequences realizing labelled perfect matchings
Joseph Briggs, Jessica McDonald, Songling Shan
https://arxiv.org/abs/2510.01110 https://arxiv.org/pdf…

Degree sequences realizing labelled perfect matchings
Let $n\in \mathbb{N}$ and $d_1 \geq d_2 \geq d_n\geq 1$ be integers. There is characterization of when $(d_1, d_1, \ldots, d_n)$ is the degree sequence of a graph containing a perfect matching, due to results of Lovász (1974) and Erdős and Gallai (1960). But \emph{which} perfect matchings can be realized in the labelled graph? Here we find the extremal answers to this question, showing that the sequence $(d_1, d_2, \ldots, d_n)$: (1) can realize a perfect matching iff it can realize $\{(1, n)…

TimeGazer: Temporal Modeling of Predictive Gaze Stabilization for AR Interaction
Yaozheng Xia, Zaiping Zhu, Bo Pang, Shaorong Wang, Sheng Li
https://arxiv.org/abs/2510.01561 htt…

TimeGazer: Temporal Modeling of Predictive Gaze Stabilization for AR Interaction
Gaze stabilization is critical for enabling fluid, accurate, and efficient interaction in immersive augmented reality (AR) environments, particularly during task-oriented visual behaviors. However, fixation sequences captured in active gaze tasks often exhibit irregular dispersion and systematic deviations from target locations, a variability primarily caused by the combined effects of human oculomotor physiology, insufficient AR headset tracking and calibration accuracy, and environmental dist…

Quantitative growth of multi-recurrence sequences
Clemens Fuchs, Armand Noubissie
https://arxiv.org/abs/2510.01896 https://arxiv.org/pdf/2510.01896

Quantitative growth of multi-recurrence sequences
In 1982, Schlickewei and Van der Poorten claimed that any multi-recurrence sequence has, essentially, maximal possible growth rate. Fourty years later, Fuchs and Heintze provided a non-effective proof of this statement. In this paper, we prove a quantitative version of that result by giving an explicit upper bound for the maximal possible growth rate of a multi-recurrence. Moreover, we also give a function field analogue of the result, answering a question posed by Fuchs and Heintze when provin…

NoTVLA: Narrowing of Dense Action Trajectories for Generalizable Robot Manipulation
Zheng Huang, Mingyu Liu, Xiaoyi Lin, Muzhi Zhu, Canyu Zhao, Zongze Du, Xiaoman Li, Yiduo Jia, Hao Zhong, Hao Chen, Chunhua Shen
https://arxiv.org/abs/2510.03895

NoTVLA: Narrowing of Dense Action Trajectories for Generalizable Robot Manipulation
Vision-Language-Action (VLA) models represent a pivotal advance in embodied intelligence, yet they confront critical barriers to real-world deployment, most notably catastrophic forgetting. This issue stems from their overreliance on continuous action sequences or action chunks, which inadvertently create isolated data silos that disrupt knowledge retention across tasks. To tackle these challenges, we propose the Narrowing of Trajectory VLA (NoTVLA) framework: a novel approach that narrows its …

Towards Understanding the Shape of Representations in Protein Language Models
Kosio Beshkov, Anders Malthe-S{\o}renssen
https://arxiv.org/abs/2509.24895 https://

Towards Understanding the Shape of Representations in Protein Language Models
While protein language models (PLMs) are one of the most promising avenues of research for future de novo protein design, the way in which they transform sequences to hidden representations, as well as the information encoded in such representations is yet to be fully understood. Several works have attempted to propose interpretability tools for PLMs, but they have focused on understanding how individual sequences are transformed by such models. Therefore, the way in which PLMs transform the wh…

Convergence analysis of inexact MBA method for constrained upper-$\mathcal{C}^2$ optimization problems
Ruyu Liu, Shaohua Pan
https://arxiv.org/abs/2511.09940 https://arxiv.org/pdf/2511.09940 https://arxiv.org/html/2511.09940
arXiv:2511.09940v1 Announce Type: new
Abstract: This paper concerns a class of constrained optimization problems in which, the objective and constraint functions are both upper-$\mathcal{C}^2$. For such nonconvex and nonsmooth optimization problems, we develop an inexact moving balls approximation (MBA) method by a workable inexactness criterion for the solving of subproblems. By leveraging a global error bound for the strongly convex program associated with parametric optimization problems, we establish the full convergence of the iterate sequence under the partial bounded multiplier property (BMP) and the Kurdyka-{\L}ojasiewicz (KL) property of the constructed potential function, and achieve the local convergence rate of the iterate and objective value sequences if the potential function satisfies the KL property of exponent $q\in[1/2,1)$. A verifiable condition is also provided to check whether the potential function satisfies the KL property of exponent $q\in[1/2,1)$ at the given critical point. To the best of our knowledge, this is the first implementable inexact MBA method with a full convergence certificate for the constrained nonconvex and nonsmooth optimization problem.
toXiv_bot_toot

Ramsey sequences with bounded clique size
Abhishek Girish Aher, Aparna Lakshmanan S
https://arxiv.org/abs/2509.23929 https://arxiv.org/pdf/2509.23929

Ramsey sequences with bounded clique size
A sequence of graphs $ \{G_k\} $ is a Ramsey sequence if for every positive integer $ k $, the graph $ G_k $ is a proper subgraph of $ G_{k+1} $, and there exists an integer $n > k$ such that every red-blue coloring of $ G_n $ contains a monochromatic copy of $ G_k $. Among the wide range of open problems in Ramsey theory, an interesting open question is ``Does there exist an ascending sequence $\{G_k\}$ with $\lim_{k \to \infty} χ(G_k) = \infty$ and $\lim_{k \to \infty} ω(G_k) \neq \infty$ t…

GUI-KV: Efficient GUI Agents via KV Cache with Spatio-Temporal Awareness
Kung-Hsiang Huang, Haoyi Qiu, Yutong Dai, Caiming Xiong, Chien-Sheng Wu
https://arxiv.org/abs/2510.00536

GUI-KV: Efficient GUI Agents via KV Cache with Spatio-Temporal Awareness
Graphical user interface (GUI) agents built on vision-language models have emerged as a promising approach to automate human-computer workflows. However, they also face the inefficiency challenge as they process long sequences of high-resolution screenshots and solving long-horizon tasks, making inference slow, costly and memory-bound. While key-value (KV) caching can mitigate this, storing the full cache is prohibitive for image-heavy contexts. Existing cache-compression methods are sub-optima…

Expressive Value Learning for Scalable Offline Reinforcement Learning
Nicolas Espinosa-Dice, Kiante Brantley, Wen Sun
https://arxiv.org/abs/2510.08218 https://

Expressive Value Learning for Scalable Offline Reinforcement Learning
Reinforcement learning (RL) is a powerful paradigm for learning to make sequences of decisions. However, RL has yet to be fully leveraged in robotics, principally due to its lack of scalability. Offline RL offers a promising avenue by training agents on large, diverse datasets, avoiding the costly real-world interactions of online RL. Scaling offline RL to increasingly complex datasets requires expressive generative models such as diffusion and flow matching. However, existing methods typically…

Adaptive User Interest Modeling via Conditioned Denoising Diffusion For Click-Through Rate Prediction
Qihang Zhao, Xiaoyang Zheng, Ben Chen, Zhongbo Sun, Chenyi Lei
https://arxiv.org/abs/2509.19876

Adaptive User Interest Modeling via Conditioned Denoising Diffusion For Click-Through Rate Prediction
User behavior sequences in search systems resemble "interest fossils", capturing genuine intent yet eroded by exposure bias, category drift, and contextual noise. Current methods predominantly follow an "identify-aggregate" paradigm, assuming sequences immutably reflect user preferences while overlooking the organic entanglement of noise and genuine interest. Moreover, they output static, context-agnostic representations, failing to adapt to dynamic intent shifts under varying Query-User-Item-C…

A note on the distribution of the sum of lengths of the initial longest increasing sequences in cycles of random permutations
Ljuben Mutafchiev
https://arxiv.org/abs/2509.26129 …

A note on the distribution of the sum of lengths of the initial longest increasing sequences in cycles of random permutations
Let $S_n$ be the set of all permutations of $\{1,2,\ldots,n\}$ and let $σ=(σ_1,σ_2,\ldots,σ_n)\in S_n$. The {\it initial longest increasing sequence} (ILIS) in $σ$ has length $m$ if, for $1\le m\le n-1$, $σ_1<σ_2<\ldots<σ_m, σ_m>σ_{m+1}$, and has length $n$ if $σ=(1,2,\ldots,n)$. Let $l(σ)$ be the length of the ILIS in $σ$. We assume that $σ$ is represented in cycle notation, so that the first number in each cycle is the minimum number of this cycle. We also assume that $σ$ is ch…

SoftCFG: Uncertainty-guided Stable Guidance for Visual autoregressive Model
Dongli Xu, Aleksei Tiulpin, Matthew B. Blaschko
https://arxiv.org/abs/2510.00996 https://

SoftCFG: Uncertainty-guided Stable Guidance for Visual autoregressive Model
Autoregressive (AR) models have emerged as powerful tools for image generation by modeling images as sequences of discrete tokens. While Classifier-Free Guidance (CFG) has been adopted to improve conditional generation, its application in AR models faces two key issues: guidance diminishing, where the conditional-unconditional gap quickly vanishes as decoding progresses, and over-guidance, where strong conditions distort visual coherence. To address these challenges, we propose SoftCFG, an unce…

Edit-Based Flow Matching for Temporal Point Processes
David L\"udke, Marten Lienen, Marcel Kollovieh, Stephan G\"unnemann
https://arxiv.org/abs/2510.06050 https://

Edit-Based Flow Matching for Temporal Point Processes
Temporal point processes (TPPs) are a fundamental tool for modeling event sequences in continuous time, but most existing approaches rely on autoregressive parameterizations that are limited by their sequential sampling. Recent non-autoregressive, diffusion-style models mitigate these issues by jointly interpolating between noise and data through event insertions and deletions in a discrete Markov chain. In this work, we generalize this perspective and introduce an Edit Flow process for TPPs th…

VersionRAG: Version-Aware Retrieval-Augmented Generation for Evolving Documents
Daniel Huwiler, Kurt Stockinger, Jonathan F\"urst
https://arxiv.org/abs/2510.08109 https://

VersionRAG: Version-Aware Retrieval-Augmented Generation for Evolving Documents
Retrieval-Augmented Generation (RAG) systems fail when documents evolve through versioning-a ubiquitous characteristic of technical documentation. Existing approaches achieve only 58-64% accuracy on version-sensitive questions, retrieving semantically similar content without temporal validity checks. We present VersionRAG, a version-aware RAG framework that explicitly models document evolution through a hierarchical graph structure capturing version sequences, content boundaries, and changes be…

Divergent Infinite Series -- Ramanujan's Initial Intuition
Mario M. Attard
https://arxiv.org/abs/2510.08652 https://arxiv.org/pdf/2510.08652

Divergent Infinite Series -- Ramanujan's Initial Intuition
This paper investigates Srinivasa Ramanujan's initial intuitive methodology for assigning the finite value -1/12 to the sum of the divergent infinite series of all positive integers. We systematically examine Ramanujan's initial method, originally sketched in his notebooks, and set the methodology into an algebraic framework. The methodology has limited applicability to other classes of divergent series. The methodology is extended to assign a Ramanujan smoothed sum to the infinite sequences of…

Discrete Facial Encoding: : A Framework for Data-driven Facial Display Discovery
Minh Tran, Maksim Siniukov, Zhangyu Jin, Mohammad Soleymani
https://arxiv.org/abs/2510.01662 htt…

Discrete Facial Encoding: : A Framework for Data-driven Facial Display Discovery
Facial expression analysis is central to understanding human behavior, yet existing coding systems such as the Facial Action Coding System (FACS) are constrained by limited coverage and costly manual annotation. In this work, we introduce Discrete Facial Encoding (DFE), an unsupervised, data-driven alternative of compact and interpretable dictionary of facial expressions from 3D mesh sequences learned through a Residual Vector Quantized Variational Autoencoder (RVQ-VAE). Our approach first extr…

Strict Log-concavity of $k$-coloured Partitions
Kathrin Bringmann, Ben Kane, Anubhab Pahari, Larry Rolen
https://arxiv.org/abs/2510.07769 https://arxiv.org…

Strict Log-concavity of $k$-coloured Partitions
In recent years, there has been extensive work on inequalities among partition functions. In particular, Nicolas, and independently DeSalvo--Pak, proved that the partition function $p(n)$ is eventually log-concave. Inspired by this and other results, Chern--Fu--Tang first conjectured log-concavity of $k$-coloured partitions. Three of the authors and Tripp later proved this conjecture by introducing recursive sequences and a strict inequality for fractional partition functions, giving explicit e…

Improving Discrete Diffusion Unmasking Policies Beyond Explicit Reference Policies
Chunsan Hong, Seonho An, Min-Soo Kim, Jong Chul Ye
https://arxiv.org/abs/2510.05725 https://…

Improving Discrete Diffusion Unmasking Policies Beyond Explicit Reference Policies
Masked diffusion models (MDMs) have recently emerged as a novel framework for language modeling. MDMs generate sentences by iteratively denoising masked sequences, filling in [MASK] tokens step by step. Although MDMs support any-order sampling, performance is highly sensitive to the choice of which position to unmask next. Prior work typically relies on rule-based schedules (e.g., max-confidence, max-margin), which provide ad hoc improvements. In contrast, we replace these heuristics with a lea…

Schur log-concavity and the quantum Pascal triangle
\'Alvaro Guti\'errez, Christian Krattenthaler
https://arxiv.org/abs/2509.22648 https://arxiv.or…

Schur log-concavity and the quantum Pascal triangle
We say a sequence $f_0, f_1, f_2, \ldots$ of symmetric functions is Schur log-concave if $f_n^2 - f_{n-1}f_{n+1}$ is Schur positive for all $n\ge1$. We conjecture that a very general class of sequences of Schur functions satisfies this property, and show it for sequences of Schur functions indexed by partitions with growing first part and column. Our findings are related to work of Lam, Postnikov and Pylyavskyy on Schur positivity, and of Butler, Sagan, and the second author on $q$-log-concavit…

IMAGEdit: Let Any Subject Transform
Fei Shen, Weihao Xu, Rui Yan, Dong Zhang, Xiangbo Shu, Jinhui Tang
https://arxiv.org/abs/2510.01186 https://arxiv.org/p…

IMAGEdit: Let Any Subject Transform
In this paper, we present IMAGEdit, a training-free framework for any number of video subject editing that manipulates the appearances of multiple designated subjects while preserving non-target regions, without finetuning or retraining. We achieve this by providing robust multimodal conditioning and precise mask sequences through a prompt-guided multimodal alignment module and a prior-based mask retargeting module. We first leverage large models' understanding and generation capabilities to pr…

Evolutionary Profiles for Protein Fitness Prediction
Jigang Fan, Xiaoran Jiao, Shengdong Lin, Zhanming Liang, Weian Mao, Chenchen Jing, Hao Chen, Chunhua Shen
https://arxiv.org/abs/2510.07286

Evolutionary Profiles for Protein Fitness Prediction
Predicting the fitness impact of mutations is central to protein engineering but constrained by limited assays relative to the size of sequence space. Protein language models (pLMs) trained with masked language modeling (MLM) exhibit strong zero-shot fitness prediction; we provide a unifying view by interpreting natural evolution as implicit reward maximization and MLM as inverse reinforcement learning (IRL), in which extant sequences act as expert demonstrations and pLM log-odds serve as fitne…

Instant4D: 4D Gaussian Splatting in Minutes
Zhanpeng Luo, Haoxi Ran, Li Lu
https://arxiv.org/abs/2510.01119 https://arxiv.org/pdf/2510.01119

Instant4D: 4D Gaussian Splatting in Minutes
Dynamic view synthesis has seen significant advances, yet reconstructing scenes from uncalibrated, casual video remains challenging due to slow optimization and complex parameter estimation. In this work, we present Instant4D, a monocular reconstruction system that leverages native 4D representation to efficiently process casual video sequences within minutes, without calibrated cameras or depth sensors. Our method begins with geometric recovery through deep visual SLAM, followed by grid prunin…

Gather-Scatter Mamba: Accelerating Propagation with Efficient State Space Model
Hyun-kyu Ko, Youbin Kim, Jihyeon Park, Dongheok Park, Gyeongjin Kang, Wonjun Cho, Hyung Yi, Eunbyung Park
https://arxiv.org/abs/2510.00862

Gather-Scatter Mamba: Accelerating Propagation with Efficient State Space Model
State Space Models (SSMs)-most notably RNNs-have historically played a central role in sequential modeling. Although attention mechanisms such as Transformers have since dominated due to their ability to model global context, their quadratic complexity and limited scalability make them less suited for long sequences. Video super-resolution (VSR) methods have traditionally relied on recurrent architectures to propagate features across frames. However, such approaches suffer from well-known issue…

Meaningless Tokens, Meaningful Gains: How Activation Shifts Enhance LLM Reasoning
Zeru Shi, Yingjia Wan, Zhenting Wang, Qifan Wang, Fan Yang, Elisa Kreiss, Ruixiang Tang
https://arxiv.org/abs/2510.01032

Meaningless Tokens, Meaningful Gains: How Activation Shifts Enhance LLM Reasoning
Motivated by the puzzling observation that inserting long sequences of meaningless tokens before the query prompt can consistently enhance LLM reasoning performance, this work analyzes the underlying mechanism driving this phenomenon and based on these insights proposes a more principled method that allows for similar performance gains. First, we find that the improvements arise from a redistribution of activations in the LLM's MLP layers, where near zero activations become less frequent while …

PAD3R: Pose-Aware Dynamic 3D Reconstruction from Casual Videos
Ting-Hsuan Liao, Haowen Liu, Yiran Xu, Songwei Ge, Gengshan Yang, Jia-Bin Huang
https://arxiv.org/abs/2509.25183 h…

PAD3R: Pose-Aware Dynamic 3D Reconstruction from Casual Videos
We present PAD3R, a method for reconstructing deformable 3D objects from casually captured, unposed monocular videos. Unlike existing approaches, PAD3R handles long video sequences featuring substantial object deformation, large-scale camera movement, and limited view coverage that typically challenge conventional systems. At its core, our approach trains a personalized, object-centric pose estimator, supervised by a pre-trained image-to-3D model. This guides the optimization of deformable 3D G…

RiskPO: Risk-based Policy Optimization via Verifiable Reward for LLM Post-Training
Tao Ren, Jinyang Jiang, Hui Yang, Wan Tian, Minhao Zou, Guanghao Li, Zishi Zhang, Qinghao Wang, Shentao Qin, Yanjun Zhao, Rui Tao, Hui Shao, Yijie Peng
https://arxiv.org/abs/2510.00911

RiskPO: Risk-based Policy Optimization via Verifiable Reward for LLM Post-Training
Reinforcement learning with verifiable reward has recently emerged as a central paradigm for post-training large language models (LLMs); however, prevailing mean-based methods, such as Group Relative Policy Optimization (GRPO), suffer from entropy collapse and limited reasoning gains. We argue that these issues stem from overemphasizing high-probability output sequences while neglecting rare but informative reasoning paths. To address these challenges, we propose Risk-based Policy Optimization …

Breaking the Euclidean Barrier: Hyperboloid-Based Biological Sequence Analysis
Sarwan Ali, Haris Mansoor, Murray Patterson
https://arxiv.org/abs/2510.01118 https://

Breaking the Euclidean Barrier: Hyperboloid-Based Biological Sequence Analysis
Genomic sequence analysis plays a crucial role in various scientific and medical domains. Traditional machine-learning approaches often struggle to capture the complex relationships and hierarchical structures of sequence data when working in high-dimensional Euclidean spaces. This limitation hinders accurate sequence classification and similarity measurement. To address these challenges, this research proposes a method to transform the feature representation of biological sequences into the hy…