跳到正文
研踪PaperTrail

论文库

检索计算机领域论文,按研究方向、作者或关键词筛选。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+479新增论文
+361其中 AI 相关
+63已有论文修订
+93新增 AI 导读

arXiv 最近完成扫描:2026-11-16 01:34 · 全站最近入库:2026-11-16 13:44各信源进度 →

统计口径与数据新鲜度

统计 2026-11-15 13:50 — 2026-11-16 13:50(北京时间) · 点击数字查看对应简报;AI 相关包含在新增论文内。

最近完成检查:2026-11-16 01:34最近发现新增版本:2026-11-16 13:44库内论文最新更新时间:2026-11-14 02:59

每分钟刷新已保存的数据。空检查不代表采集异常;新收录可能包含历史论文。AI 导读按新增版本计数。

23,827 已收录论文19,941 AI 相关6,921 有修订112 HF 社区发现

浏览论文库

时间范围跟随排序:版本更新、本站收录或首次发表。最新收录可能包含历史论文。部分信源异常,查看状态 →

找到 1,173 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-16 01:33(北京时间)
新论文 · v1arXiv:2611.16267

Causals: An Open Knowledge Graph for Research on Causal Reasoning

Miaozhe Lin, Qiongzheng Lin, Zhongwei Liang, Zijun Liu

原文摘要

Applying causal knowledge to a new event requires connecting the event to relevant entities and the relationships through which its consequences may unfold. We present Causals, an open knowledge graph for research on this form of reasoning. Its initial release, constructed from news reports, contains 369,184 nodes and 1,272,418 edges linking reported events, organizations, commodities, financial instruments, and other entities. Structural relationships, observations, and attributed causal claims retain distinct semantics and links to source records. We describe the construction and representation of the resource and characterize its connectivity and multi-hop retrieval properties. Causals provides shared relational knowledge for studying how language models retrieve relevant context and apply mechanisms to particular events. The dataset is available under CC BY 4.0 at https://causals.org/, and we invite contributions to its ontology, records, and evidence.

cs.AI

新论文 · v1arXiv:2611.16260

Low Language Transfer Speech-to-Text

Husein Zolkepli

原文摘要

A multilingual speech recogniser has heard much speech in its own languages. We ask how much of that it brings to a language it was never trained on, when it is fine-tuned on the little that language has. We score 28 open multilingual models untouched on Low-Language-TTS, 50 low-resource languages in 11 writing systems, and fine-tune 24 of them on at most 1,000 clips a language, 82 hours in all, with a dev set choosing every epoch. Fine-tuning lowers every model's macro character error rate (CER), by 16.6 to 53.1 points, to 9.5 at best, and a word language model takes three CTC models 1.2 to 1.4 points lower; WER is still 34.4 at best. On the 12 languages that no model lists, every model starts at 46 or worse and the best ends at 11.7. The same architectures from random weights stay 60 points or more behind, so the gain is transfer, and an encoder left untouched already gives a Whisper or Qwen decoder most of what it needs. It is the language, not the corpus: on FLEURS recordings that ours do not share, the median model keeps 79% of its gain. It costs the languages the models knew, 5.7 points of CER on English, Spanish, French and German in the median model. Low-rank adapters on 4% of the weights come within 1.1 points of tuning everything, and on 2 of 3 models forget less. Seq2seq and language-model decoders are done within three to five epochs, while a new CTC head is still improving at twenty or thirty. Data beats epochs: 25 clips a language take Qwen3-ASR 0.6B from 60.9 to 39.3, and at equal clips presented, ten times the data is worth 9.7 points. A model that first learned forty of the languages learns the other ten no faster. On a public Malay benchmark, untouched Whisper large-v3 scores 8.9 when it detects the language and 14.9 when told the clip is Malay, because it then translates the English that speakers mix in; trained on the Scicom telephony dataset it scores 6.0.

cs.CL

版本更新 · v2arXiv:2605.07979

The Limits of AI-Driven Allocation: Optimal Screening under Aleatoric Uncertainty

Santiago Cortes-Gomez, Mateo Dulce Rubio, Carlos Patino, Bryan Wilder

原文摘要

The rise of machine learning has shifted targeted resource allocation in policy and humanitarian settings toward algorithmic targeting based on predicted risk scores. This approach is typically cheaper and faster than traditional screening procedures that directly observe the latent vulnerability status through physical verification. Yet, even access to the true conditional vulnerability probability cannot eliminate misallocation: aleatoric uncertainty over individual vulnerability status is irreducible, and probabilistic targeting inevitably misallocates some resources. In this work, we study how screening and algorithmic targeting should be optimally combined in a two-stage allocation framework, in which a screening stage observes true outcomes for a subset of units before a final allocation stage assigns the resource under a fixed coverage budget. We show that the optimal strategy screens units at the margin of algorithmic allocation, while directly targeting the highest-risk units. We empirically characterize when screening and algorithmic targeting act as complements or substitutes: efficiency gains from screening grow as the aleatoric uncertainty in the population increases. Finally, we characterize the relative value of screening and direct allocation, which yields a simple rule for splitting a program budget between the two: the optimal cutoff depends only on their cost ratio. We illustrate our framework with applications in income-based social protection programs and humanitarian demining in Colombia, where the tension between screening costs and allocation efficiency is operationally consequential

cs.AI

新论文 · v1arXiv:2611.16155

Accelerating Distributed Variational Quantum Computing over Wireless Networks with Incentive Mechanism Design

Shaba Shaon, Dinh C. Nguyen, Octavia Dobre

原文摘要

Distributed variational quantum computing (DVQC) appears as the emerging paradigm to address the limitations of near-term quantum devices by decomposing large parameterized quantum circuits (PQCs) into smaller subcircuits executed across multiple quantum processors. In this context, how to incentivize quantum processors to participate actively is crucial for accelerating DVQC, yet remains largely unexplored in the literature. To address this significant gap, we propose a novel two-stage incentive framework for wireless DVQC. In the first stage, we minimize processor-side energy cost via a joint optimization of subcircuit parameters, number of noise measurements, and transmit power, efficiently solved using block coordinate descent (BCD) and successive convex approximation (SCA). In the second stage, we maximize aggregator-side system utility by selecting an optimal subset of processors based on their optimized energy costs derived from the first stage, formulating a non-convex combinatorial selection problem addressed using the alternating direction method of multipliers (ADMM). Extensive numerical experiments on the MNIST and CIFAR-10 datasets demonstrate that our DVQC framework achieves training and testing performance comparable to centralized quantum computing in terms of accuracy and loss, with similar convergence behavior on MNIST and slightly slower initial convergence on CIFAR-10. Moreover, our optimization approach consistently outperforms baseline schemes in energy efficiency and system utility, revealing a strong synergy between processor-level energy minimization and incentive-driven processor selection, where reduced energy costs directly enhances a processor's likelihood of selection and contributes to higher overall system utility.

cs.DC

新论文 · v1arXiv:2611.16048

Privacy Leakage from Recurrent State in LLM Inference

Weimin Fu, Shikun Liu, Qiying Li, Yuntao Liu, Pan Li, Xiaolong Guo

原文摘要

Personalized LLM services place private facts and application instructions into a prompt, then retain representations of that context for later inference. Latent-state LLMs compress part of this history into fixed-size recurrent state. The information preserved by its learned updates is implicit, and later computation can expose it even when an observer receives neither the private text nor the responses. We investigate privacy leakage from recurrent state through partial-state probing and physical observation of its use. Our methodology separates recoverable information, persistence through public dialogue, causal influence on later decisions, and leakage through power. Experiments with released hybrid implementations show that a small state observation supports multiple predictions about private facts and response rules, and that personal attributes generalize to unseen authors. The information remains inferable after public dialogue and changes later decision margins. Complete-read power recovers private facts under a fixed execution schedule. On a partitioned GPU, an unprivileged account infers age from whole-device power during public requests; a separate state-isolation experiment establishes the recurrent contribution to this channel. These findings extend context confidentiality to the lifetime and execution of retained state. They identify access-control responsibilities for serving systems, telemetry restrictions for GPU operators, and data-dependent power leakage for accelerator designers.

cs.CR

版本更新 · v2arXiv:2610.29037

Near-Optimal Consistency-Robustness Bounds for Non-Clairvoyant Scheduling

Sizhe Li, Nicolas Christianson, Tongxin Li

原文摘要

We study the problem of characterizing the optimal consistency--robustness tradeoff for single-machine non-clairvoyant scheduling with predicted processing times. For a fixed number of jobs $n$, Wei and Zhang (2020) established a lower bound on the achievable consistency--robustness tradeoff which is known to be tight only in a few restrictive cases, leaving the optimal tradeoff for general $n$ unresolved. A different worst-case-$n$ regime was introduced by Benomar et al. (2025), who applied the former work's construction to show that for $λ\in (0, 1)$, any $(1+λ)$-consistent algorithm must be $1 + Ω(\sqrt{1/λ})$-robust. On the other hand, the best known upper bound gives $(1+λ)$-consistency and $1+O(1/λ)$-robustness, leaving a polynomial gap. In this work, we develop a new lower bound using a multi-stage adversarial construction and a new algorithm that together give a near-optimal characterization of the consistency--robustness tradeoff: for every $n$, our upper bound is within a multiplicative factor $3/2$ of our lower bound, and thus our algorithm is within a multiplicative factor $3/2$ of the online optimal. In particular, our lower bound shows that any $(1+λ)$-consistent algorithm must have robustness at least $1/\sqrtλ$ as $n \to \infty$, while our algorithm achieves robustness at most $\frac{1}{2}+\sqrt{\frac{5}{4}+1/λ}$ for any $n$, matching the lower bound in the leading term as $λ\downarrow 0$; our upper bound also closes the polynomial gap in the worst-case-$n$ regime. We additionally show how our algorithm can be made error-tolerant, preserving the near-optimal consistency-robustness tradeoff while incurring only slightly degraded performance under bounded prediction errors.

cs.DS

新论文 · v1arXiv:2611.16000

Predicting the Needle in a Petabyte Scale Haystack: Open-Vocabulary Event Anticipation in Satellite Imagery

Shreelekha Revankar, Mikhail Klassen, Creon Levit, Ash Hoover, Kavita Bala, Bharath Hariharan

原文摘要

Satellite constellations now image the Earth daily, capturing the lifecycle of events from deforestation to urban construction. Anticipating these events before completion enables timely intervention, yet existing systems cannot jointly identify where a change occurs, what it will become, and when it will finish across an open vocabulary. Building such a model requires diverse event data, but events are like needles in a haystack, making manual annotation at global scale infeasible. We address this problem by representing change as the vector difference between vision-language model embeddings at distinct time points. This arithmetic approach allows us to search for semantic transformations directly in the latent space, enabling (1) an automated data engine for global event discovery without manual annotation and (2) the largest global satellite event dataset to our knowledge, comprising 21,000+ locations across 64+ event types with ~2M images from PlanetScope and Sentinel-2 spanning six continents, to train (3) a novel global-scale event anticipation model. Our model detects change with a 93.5% max F1 score, outperforms baselines 3.6x on predicted event retrieval, and forecasts completion dates with a 3-day median error, even nine months before completion. Code: https://github.com/ShreelekhaR/predicting-the-needle

cs.CV · cs.IR

版本更新 · v2arXiv:2611.14284

Ninety-Six Ways to Measure a Scribe: Fragmented Evaluation of Ambient Scribes in Healthcare

Amama Mahmood, Shiye Cao, Laura Fleig, Junan Jiang, Theodore Jensen, Yee-Yin Choong, Kristen Greene, Chien-Ming Huang

原文摘要

Ambient artificial intelligence (AI) scribes have moved rapidly from experimental pilots to enterprise-wide deployment across health systems, promising reduced documentation burden and improved clinician experience. While research evaluating these tools has expanded quickly, there are no agreed-upon protocols or standardized metrics for assessing their performance, making it difficult to compare findings or cumulatively synthesize evidence across studies. Across 54 recent evaluation studies, we identify ten evaluation dimensions and 40 distinct metrics, operationalized in 96 different ways-35 of them (36%) used by only a single study. Assessment is dominated by documentation efficiency and time, measured in 44 of 54 studies. The other dimensions appear at broadly similar rates (16-24 studies each), and safety and clinical outcome least consistently of all. Ripple effects on the broader care system, such as impact on patients' care outcomes, redistribution of work, changes in billing and coding intensity, and downstream use of notes by other clinicians are rarely assessed. Moreover, the metrics used rely heavily on self-report and are fragmented and inconsistently operationalized, which limits comparability and hinders synthesis across studies. As an effort to support more rigorous and cumulative evaluation, in this paper we review the study designs, data sources, and metrics commonly used to evaluate ambient AI scribes in healthcare settings, discuss the limitations of current evaluation practice, and argue for a more holistic approach grounded in a standardized, multi-dimensional set of metrics.

cs.HC · cs.AI

新论文 · v1arXiv:2611.15952

Quasi-Cyclic Qudit Lifted Product Codes with Almost Linear Minimum Distance

Shantom K. Borah, Abhinav Vaishya, V. Lalitha, Asit Kumar Pradhan, Bane Vasic, Narayanan Rengaswamy

原文摘要

Quantum low-density parity-check (QLDPC) codes are a promising route to fault-tolerant quantum computation with low overhead, and qudits offer advantages over qubits including more efficient information representation, improved magic state distillation, and richer algebraic structure for code design. Asymptotically good QLDPC codes arise from the lifted product (LP) construction with non-abelian lifts, but cyclic lifts remain attractive in practice because the resulting quasi-cyclic codes have circulant structure that hardware decoders exploit, which matters especially for non-binary decoders with costly check node operations. We introduce a qudit generalization of the LP construction over cyclic group algebras, which extends hypergraph product and LP codes to arbitrary prime-power dimensions and yields quasi-cyclic qudit CSS codes for every stacking order. We bound encoding rates via a connection to additive codes, and show that in the fully stacked case, qudit LP codes built from lifted Tanner codes on expander graphs form families with $Θ(\log N)$ logical qudits and almost linear minimum distance $Θ(N/\log N)$, via a distance bound that holds for a general class of LP factors. We also study an additive decoder for general qudit codes and a linear merged-syndrome decoder for fully stacked codes. Monte Carlo simulations under a depolarizing channel with equal symbol error probability show that qudit codes achieve lower frame error rates than comparable qubit codes, and that the linear decoder gives further gains for fully stacked codes.

quant-ph · cs.IT

新论文 · v1arXiv:2611.15924

The Distributed Run Manager A C++ Library for Distributed Processing

Brian Harrawood

原文摘要

The DistributedRunManager (DRM) is a C++ library for distributing a computing task among a number of computer systems for parallel processing on a network. The library was originally designed to accelerate synthetic image production for tomographic images, but is also a general solution for distributing a software task across diverse, multiple hardware platforms. The DRM self-configures using a novel C++ class, the ElectionManager. This class selects a server cooperatively at runtime and eliminates the need for a priori distribution configuration. All communication, logging, and fault-tolerance mechanisms are encapsulated by the DRM, providing a "black-box" interface to the application developer.

cs.CE

新论文 · v1arXiv:2611.15837

The BEATS Dataset: Towards Understanding Music-Physiology Dynamics in Real-World Scenarios

Benjamin Gutierrez Serafin, Tanaya Guha, Fahim Kawsar

原文摘要

We introduce the BEATS dataset, a robust compilation of psycho-physiological data collected in the wild from 30 participants while they listened to curated playlists intended to promote well-being. An acronym for Biomarkers, Experience and Affect from Therapeutic Soundscapes, the BEATS dataset aims to support research on wellbeing, music psychology, affective computing and signal processing. More specifically, this dataset includes: (1) sensor data collected during music listening sessions using a research-grade wearable device in free-movement environments; (2) music listening activity records, stimulus metadata, acoustic features, and self-reports associated with the listening sessions (3) individual characteristics, such as demographics, music preference, musical skills, and personality traits; and (4) self-reported psychological measures before and after their experimental participation spanning for a number of weeks. By situating data collection outside of lab settings, the BEATS dataset offers the opportunity to study real-world music-listening scenarios, allowing for a richer exploration of the physiological, subjective and therapeutic effects of music under real-world conditions.

eess.SP · cs.AI

新论文 · v1arXiv:2611.15817

Tree Decomposition of Product and Nest Graphs for Assortment Optimization under the Generalized Nested Logit Model

Yuheng Wang, Paat Rusmevichientong, Huseyin Topaloglu

原文摘要

The generalized nested logit model extends the nested logit model by allowing each product to belong to multiple nests. The overlapping nests in the generalized nested logit model can capture complex substitution patterns, and the model has found applications in many areas. We study the assortment optimization problem under the generalized nested logit model and prove that it is NP-hard when there are only two nests or when any two distinct nests have at most one product in common. Motivated by these complexity results, we introduce novel concepts of product and nest graphs that capture which products are included in which nests. We use the tree decompositions of the graphs to characterize when the assortment optimization problem admits efficient solution methods. When the treewidth of a product graph is bounded, the assortment optimization problem can be solved in polynomial time. If the treewidth of the product graph is unbounded but the treewidth of the nest graph is bounded, then there exists a fully polynomial-time approximation scheme. These results represent the first polynomial-time algorithms for assortment optimization under the generalized nested logit model. Our numerical experiments on synthetic and real datasets show that the generalized nested logit model provides a better fit to the data than many existing models in the literature.

cs.DS · math.CO

新论文 · v1arXiv:2611.15803

Statistical-to-computational gaps in central limit theorems for high-dimensional tensors

Debsurya De, Dmitriy Kunisky

原文摘要

We study high-dimensional central limit theorems (CLTs) as hypothesis testing problems, comparing the number of summands needed for statistical and computational indistinguishability of the normalized sum from the limiting Gaussian distribution as the dimension and number of summands grow together. For the off-diagonal entries of sums of i.i.d. rank-one tensors formed from standard Gaussian vectors (equivalently, the problem of distinguishing Wigner from Wishart tensors), we prove total variation bounds sharpening previous estimates of Mikulincer (2020) and give matching polynomial-time tests. This shows that the statistical and computational thresholds coincide for this CLT, contrary to a proposal by Kunisky, Moore, and Wein (2024). On the other hand, we also show that, for tensors of order at least three, sufficient Gaussian dilution of the summands by adding independent Gaussian noise and rescaling gives a CLT model with a statistical-to-computational gap (in the above setting and ones with more general rank-one summand distributions). To the best of our knowledge, such a phenomenon has not been observed before in any high-dimensional CLT. In particular, in the relevant parameter regimes, we show that a brute-force search achieves strong detection, while polynomials of degree up to $n^{Ω(1)}$ cannot achieve strong separation. The latter lower bounds control low-degree $χ^2$-divergences by a novel method based on cumulant generating functions, and we further illustrate the use of these tools by applying them to masked tensor CLT models where only a prescribed subset of tensor entries is revealed.

math.ST · cs.CC · math.PR

版本更新 · v3arXiv:2609.13630

Minimum central circles: an effective characterization of the global asymptotic constant

Maurizio Falconi

原文摘要

Let $R^\ast(n)$ be the minimum radius of a central circle externally tangent to nonoverlapping circles of radii $1,\ldots,n$. We prove that $R^\ast(n)=C_\ast n^2+o(n^2)$ and give finite linear programs that enclose $C_\ast$ with prescribed error. An exact angular comparison reduces the problem to marked points on a line with separation at least the geometric mean of each pair of marks. Concatenation with bounded joining cost and reassignment to the prescribed normalized radii prove existence of the limit. Balanced finite-word programs give a bracket of width $(1/k+1/r)/π$, before directed arithmetic error, using $k$ types and words of length $r$. We also obtain explicit geometric bounds. For the upper bound, we recover finite and countable reflected-block constructions by actual permutations, retaining all pair constraints. A coupled terminal-subset argument gives an explicit lower bound strictly above $1/8$. Together these give $C_{\rm term}+η_{\rm width}\le C_\ast\le U_4$. The lower endpoint comes from the terminal-subset argument, and the upper endpoint from the four-block construction; neither is asserted sharp. The finite-word scheme establishes computability; efficient high-precision evaluation and the structure of global optimizers remain open.

math.MG · cs.CG

版本更新 · v3arXiv:2601.19070

Critical Organization of p-Adic Deep Neural Networks

W. A. Zúñiga-Galindo

原文摘要

We study continuous neural networks whose neurons are the points of the ring of p-adic integers, a fractal tree-like space that encodes hierarchical architectures. The hidden states solve a nonlinear integral equation, which describes the equilibria of recurrent networks and deep equilibrium models; feedforward architectures appear as the special case of kernels without feedback. We prove that the state is unique and depends Lipschitz-continuously on the parameters whenever the product of the Lipschitz constant of the activation function and the operator norm of the connectivity is less than one, and that, when the parameters are locally constant, every state is exactly the state of a finite network with a tree-like topology. We introduce p-adic discrete deep networks, whose layers are the levels of the tree and whose weights are coarse-grainings of a single kernel, and show that, in the stable region, they converge to the continuous network as the depth tends to infinity. Feedforward networks always have a unique state; hence a multiplicity of states, and therefore critical organization in our sense, requires recurrence. For recurrent networks with radial kernels we prove a sharp transition: the zero state is the only state if and only if the total coupling does not exceed one. For hierarchical kernels, above the threshold, the states are indexed by sign patterns on the leaves of the tree, their number is computed exactly, and they form a tree-indexed partially ordered set graded by instability. Numerical experiments illustrate these results and test conjectures on general radial kernels and on random recurrent networks on the tree. Finally, we interpret the critical organization hypothesis for deep neural networks, originally formulated for brain organization, through the lens of multistability, and we discuss the continuous limits of families of finite networks.

cs.LG

新论文 · v1arXiv:2611.15698

Learning Mixtures of Exponential Family

Mansi Sood, Devavrat Shah

原文摘要

Motivated by causal inference and physical computing systems, we consider the problem of learning mixture of two bounded-support exponential-family distributions with a shared interaction parameter matrix and mixture component-specific linear parameter vectors. The unknown sequence of component labels belongs to a known structured class. We propose an estimator based on a convex loss that avoids the log-partition function; for each fixed label assignment, estimation is a convex optimization problem. We show that the conditional population loss uniquely identifies the interaction matrix and the linear parameter attached to each sample. For a global empirical minimizer, we establish finite-sample bounds on the interaction error in Frobenius norm and the average squared linear-parameter error. The bounds depend on the logarithm of the number of admissible label assignments, and, for fixed dimension and model bounds, imply consistency when this complexity grows sublinearly with sample size. Additional occupancy and separation conditions yield component and label recovery.

stat.ML · cs.LG · eess.SP

新论文 · v1arXiv:2611.15681

Energy Aware Optimization of Non-Deterministic Task-Based Intermittent Computing Nodes with Thimble

John Madden, Morgan Masters, Alec Levy, Steve McGuire, Colleen Josephson

原文摘要

Intermittent computing nodes run on harvested energy, which they store in capacitors and spend in short bursts of computation. Choosing those capacitors is a trade-off: large capacitors can power large tasks but leak more current, wasting harvested energy, while small capacitors leak less but cannot power large tasks. Recent work, LEACS, showed that selecting capacitors and mapping tasks to them at compile time improves inference throughput. However, LEACS optimizes the mapping for a predetermined workload, whereas sensor nodes often run data-dependent workloads in which the next task depends on the outcome of earlier ones. We present Thimble, an energy storage system that maps a node's tasks to capacitors at runtime. Thimble models the workload as a Markov chain over tasks and assigns each task to a capacitor by weighing capacitor leakage, the task's energy demand and, when it replans, the likelihood of each future task. We compare Thimble against a reimplementation of LEACS, which ranks every static plan by simulation. When the training trace resembles deployment, Thimble's single design-time solve comes within 0.3--4% of LEACS's throughput. When the deployment conditions drift from the characteristics of the training set, LEACS's throughput-focused static plans can fail sharply. On the lowest-power source we consider, LEACS's plan stalls on deployment traces that harvest less than its training trace, and on some never wakes; Thimble avoids this, with 22--29% higher throughput at the reference workload. A sensitivity sweep over 25 workload profiles on that source shows that Thimble with depth replanning leads LEACS by 22--42% throughout, and what run-time remapping adds: replanning at a fixed interval matches or improves on the static map at 21 of 25 points, by up to 7%.

cs.NI

新论文 · v1arXiv:2611.15671

The Noise Floor of a Quantization Comparison: Engine, restart, batch and kernel drift, measured against 8-bit and 4-bit effects in two decision models

Aleks Blagojevich, Trevor Sunderland

原文摘要

A quantized model is usually judged by comparing it with its BF16 original, but that comparison changes more than the weights. It often changes the serving engine, it always runs on a different server instance, it may change how requests are batched, and on some stacks the library selects a different kernel depending only on which packages import. We measure each of these sources on the same 2,400 public decisions for two of Cloudflare's decision models, Clef-flash (9B) and Clef (27B), on an NVIDIA GB10, and state every quantization effect as the excess over them. Sources that are not quantization move between 0 and 0.92% of decisions. llama.cpp restarts are bit-identical on every record for both models. vLLM restarts flip 0, 5 and 20 decisions at BF16, FP8 and NVFP4 on Clef-flash and 4, 0 and 22 on Clef; a third instance of Clef's quantized builds (exploratory) leaves FP8 at 0 and puts NVFP4 at 4 to 22, depending on which two instances are compared. An engine change at BF16 flips 6 to 16, a batch-size change 8 to 15, and the Gated DeltaNet kernel transformers selects unasked 10 to 14. Quantization inside one engine moves more: for Clef-flash and Clef, 0.88% and 1.04% at Q8_0, 2.12% and 1.62% at FP8, 5.79% and 3.50% at Q4_K_M, and 5.79% and 5.67% at NVFP4. On both models, every 4-bit interval excludes every non-quantization source. Two of our five pre-registered predictions failed, and the same two fail when the tests are rerun on the 27B: FP8 is not inside restart noise, and changing engine moves fewer decisions than 8-bit quantization does. Accuracy hides most of this. Over the full 31,275 public questions, Clef's Q4_K_M build moves accuracy by a net six questions (-0.02 points [-0.20, +0.16]) while 436 of its decisions get worse (exploratory), and NVFP4 costs Clef-flash 0.95 points and Clef 0.65. We release the code, quantization recipes, per-record outputs and the pre-registration.

cs.LG

新论文 · v1arXiv:2611.15642

Differential Backpack Language Models: Sense Inhibition via Signed Contextualization Weights

Konstantinos Skianis, Giannis Nikolentzos, Polykarpos Meladianos, Michalis Vazirgiannis

原文摘要

Backpack language models predict with a strictly linear combination of non-contextual per-word sense vectors, weighted by non-negative contextualization scores. We show that non-negativity conflates two properties: the linear readout, which carries the interpretability, and the conic combination, which merely restricts it. We relax the latter by computing the sense weights with differential attention, a difference of two softmax maps with a learned per-sense gain, so context can actively inhibit a sense; we call the result a differential Backpack. At equal parameters on WikiText-103 this cuts the Backpack's perplexity gap to a Transformer by 31%, improves over the softmax Backpack at every sense count, persists with a shrinking margin up to 350M parameters on OpenWebText, and is localized at the sense layer by ablations and controls. The learned inhibition is directly interpretable (a financial context drives the river sense of bank to -0.50), and the preserved per-sense attribution predicts counterfactual removals exactly. We tested a number of combination rules, showing the linear interface is nearly free. Making the readout nonlinear gains only 0.43 perplexity points while invalidating two thirds of the decomposition. The differential Backpack's senses are also better representations, improving zero-shot sentence similarity by up to 33%.

cs.CL

版本更新 · v3arXiv:2606.11270

Quantifying Subliminal Behavioral Transfer Ratios in Language Model Distillation

Uwe König, Hamza Kazmi, Ruizhe Li, Maheep Chaudhary

原文摘要

Distillation of a language model intended to transfer benign behavior to a student model may also transfer undesirable characteristics, if they are present in the teacher model, a phenomenon known as subliminal learning. While qualitative evidence supports the existence of this effect, its magnitude has not been systematically characterized. This study quantifies subliminal behavioral transfer ratios by steering two teacher models (Llama-2-7B-Chat and Qwen2.5-7B-Instruct) at varying steering strengths and distilling student models using only benign data. Evaluation on 100 JailbreakBench prompts with GPT-4.1, serving as the evaluator, indicates that transfer is robust but exhibits distinct scaling behaviors. Llama-2 demonstrates a sharp threshold ($τ= {0.25,0.32} \ \text{beyond} \ α= -0.15$), whereas Qwen2.5 displays continuous and higher levels of transfer ($τ$ up to $0.61$).

cs.LG · cs.AI · cs.CL

新论文 · v1arXiv:2611.15607

FADO: Learning Causal Explanations with Prior-Fitted Networks

Seth Flaxman, Jess Carr, Annabel Jakob, Marek Masiak, Dino Sejdinovic

原文摘要

In regression and classification tasks with tabular datasets, most of the literature on explanation in ML (e.g. SHAP, permutation importance) has focused on predictive attribution: does feature $X_i$ help predict outcome $Y$? At its heart, predictive attribution captures correlation, not causation. We instead ask a causal question: would manipulating feature $X_i$ change outcome $Y$? We define a natural quantity to answer this question, the feature intervention effect: $Δ_i = E[Y\mid do(X_i=+1)]-E[Y\mid do(X_i=-1)]$. In general, estimating this quantity from observational data alone would entail strong causal assumptions. We propose FADO (Feature-wise Amortized Do-Operators), an amortized inference method using prior-fitted networks trained on synthetic tabular datasets generated by structural causal models. During training, the synthetic data-generating process is known, so we can compute ground-truth intervention effects $Δ_i$, and train a model to predict them at test time from a purely observational tabular dataset. As we demonstrate through extensive experiments, our new causal explainability method is easy to train and provides better explanations under a range of settings for which standard predictive attribution fails, including proxy variables, confounders, colliders, and placebos. We evaluate it on real, synthetic and semi-synthetic datasets, with and without hidden confounders, and compare it to existing predictive attribution methods and state-of-the-art causal inference methods.

cs.LG · stat.ML

版本更新 · v2arXiv:2511.19757

What does it mean to understand language?

Colton Casto, Anna Ivanova, Evelina Fedorenko, Nancy Kanwisher

原文摘要

Language understanding entails not just extracting the surface-level meaning of the linguistic input, but constructing rich mental models of the situation it describes. Here we propose that because processing within the brain's core language system is fundamentally limited, deeply understanding language requires exporting information from the language system to other brain regions that construct mental models of minds, objects, and places, and that compute perceptual and motor representations. Building on earlier theorizing and empirical work, we argue that recent progress in cognitive neuroscience provides both the conceptual foundation and the methods to directly evaluate this hypothesis, thus opening up a new strategy to reveal what it means, cognitively and neurally, to understand language.

cs.CL

版本更新 · v3arXiv:2609.02753

The Import Tax: A Longitudinal Measurement of Startup Cost in the Python Ecosystem

Trinath Sai Subhash Reddy Pittala

原文摘要

Python programs pay for their imports at every process start, a cost that is invisible in steady-state benchmarks but dominant for command-line tools, test workers, and serverless cold starts. Python 3.15 adds explicit lazy imports (PEP 810) on largely anecdotal evidence. We measure the import cost of the 500 most-downloaded PyPI packages, every quarterly release since mid-2021, under six CPython versions (3.9-3.14) on two platforms (Apple M5/macOS and Intel Xeon/Linux), 66,816 cells in all, with targeted experiments that isolate individual effects. Values are for macOS (Linux in parentheses). Import cost is heavily skewed (median 12.3 ms, 99th percentile 400 ms). Among packages that import in more than 2 ms, namespace distributions excluded, the median package spends 68% of its import time in the standard library (68% on Linux with the interpreter on local storage, 79% on network storage). Precompiling bytecode removes almost all of the cold-start penalty on Linux and 92% of it for the median package on macOS, but only a third of the summed macOS penalty: packages with native libraries pay a security assessment of each library that the system has not seen before, again in every fresh environment. A package's own import code grows by a median 4.0% (5.0%) per year; installing each release with the dependencies of its time changes its trend by a median of only 0.15 (0.35) percentage points (CPython 3.9). CPython 3.14 runs package code faster than 3.9 but imports more and slower standard-library modules, a net 1.16x (1.07x) slowdown. Submodules that dependents import beyond the top level add nothing for the median package and at most 11.7 (49.5) ms for nine packages in ten. On CPython 3.15, global lazy mode cuts the imports that dependents perform to 0.41x (0.39x) of their cost, and 38 of 461 packages fail under it, many only on first use of a name. Harness, code, and dataset are public.

cs.SE

版本更新 · v4arXiv:2605.02171

QuIVer: Efficient Graph Indexing with Two-Bit Construction and Navigation

Wenxuan Xiao, Peidong Zhu, Chengcheng Li, Xu Cao

原文摘要

High-dimensional vector indices must balance construction time, memory, and query throughput. Two-bit comparisons reduce computation and storage, but preserving useful graph connections and search candidates requires scores suited to each stage. We present QuIVer, a graph index that uses a training-free sign-magnitude representation for construction and navigation while reserving original vectors for bounded final reranking. Construction discovers candidates with Hamming scores, orders them with an integer proxy, and defers reverse-edge pruning across parallel insertions. Search uses reconstruction norms and bitwise inner products to order both the frontier and retained candidates, without query calibration or an exact anchor. Experiments cover 21 datasets and compare 30 external baseline modes on three core workloads. QuIVer's validation-selected 95% configurations lie on the measured Pareto frontiers of achieved recall, throughput, index preparation time, and memory, whether memory counts navigation state alone or also the original vectors. On Cohere-5M, QuIVer reaches 95.90% Recall@10 at 3,075 single-thread queries/s, $11.74\times$ the throughput of Lucene BBQ with FP32 reranking at 95.83% recall. Against SymphonyQG at 95.87%, it retains 81% of single-thread and 97% of 32-thread throughput while reducing index preparation time by 54%, total logical memory by 47%, and navigation storage by 87%. An SSD mode with INT8 reranking serves Cohere-5M at about 5 GiB of steady-state process RSS. Controlled crossovers and candidate-coverage diagnostics connect these results to construction and retention quality. Across 21 datasets, high recall holds for most text and image embeddings, while non-negative descriptors, cross-modal queries, and random data expose the limits of the representation.

cs.DB

新论文 · v1arXiv:2611.15514

Real Pairs Are Not Enough: Physically Anchored Supervision for Volumetric CT Super-Resolution

Haoran Liu, Zipei Yan, Mingzhe Liu, Jizhou Li, Guibin Zan

原文摘要

Volumetric CT super-resolution (SR) models are commonly trained on synthetic or physically acquired pairs of low-resolution (LR) and high-resolution (HR) volumes. However, synthetic pairs miss real acquisition effects, whereas registered physical pairs retain inter-scan noise, artifact, and microstructure differences---causing direct voxelwise regression, and even naive synthetic-real mixing, to oversmooth fine porous structures. We introduce the Physically-anchored Implicit Network (PIN) with Real-Pair Dual-Anchor learning, which decouples physical-pair supervision into two complementary roles. The HR anchor teaches fine structure from registered HR through pair-conditioned spectral replay, while the LR anchor constrains the prediction from acquired LR via affine-calibrated low-pass projection. Together with exact synthetic supervision in one shared volumetric implicit model, PIN learns from both physical scans without forcing the acquired-LR reconstruction to match registered HR voxel by voxel. Inference requires only acquired LR. Across 30 external baseline configurations on real CT, PIN achieves the lowest full-volume LPIPS, reduces average symmetric surface distance by 42.3-49.5% and bone-volume-fraction error by 52.2-53.4% relative to the three strongest baselines for each metric, and lowers the erased reference-pore volume fraction from 37.3-72.9% to 10.4%. Code, pretrained models, and evaluation scripts will be publicly available.

cs.CV · eess.IV

新论文 · v1arXiv:2611.15465

On-Edge Acoustic Continual Anomaly Detection: A Cross-Platform Comparative Study

Rami Zewail, Yehia Ehab, Karim Ali, Ahmed Mahrous, Mohamed Awad, Bassem Mokhtar

原文摘要

Acoustic Anomaly Detection (AAD) serves as a non-contact approach for machine condition monitoring, yet much of the existing research is evaluated offline and rarely addresses embedded hardware constraints or microphone variations under operational drift. This paper presents a comparative evaluation of on-edge continual AAD across two embedded platforms the ESP32-S3 (with an external I2S digital microphone) and the Arduino Nicla Vision (with an onboard Pulse Density Modulation (PDM) microphone) using a shared detection model. The model utilizes a lightweight ShallowNet encoder and contrastive prototype learning to map spectrograms into a 32-dimensional embedding space. Both boards run an identical TensorFlow Lite for Microcontrollers (TFLM) encoder, performing on-device prototype learning updated via an exponential moving average. A cloud pipeline orchestrated through Firebase and MQTT enables remote data collection and retraining. Evaluated via a controlled speaker-replay protocol using standardized Detection and Classification of Acoustic Scenes and Events (DCASE) Challenge Task 2 machine sounds, the ESP32-S3 achieved 82.35% to 95.00% accuracy across five machine types, while the Nicla Vision achieved 79.41% to 95.00%. The two platforms demonstrated comparable performance, confirming that the lightweight prototypical architecture transfers reliably across distinct hardware and sensor tiers.

cs.SD · cs.LG

版本更新 · v2arXiv:2611.12144

DisasterVLG: Vision-Language Grounding and Evidence-Based Reasoning for Disaster Scene Understanding

Ali Diba, Aisha Al-Mohannadi, Yin Yang, Muhammad Imran, Ferda Ofli

原文摘要

General-purpose vision-language models struggle with crisis imagery due to its inherent noise, ambiguity, and high variability across disaster events - yet timely, accurate scene understanding is critical for humanitarian response. To address this, we introduce DisasterVLG, a large-scale multi-task dataset containing 1.24 million disaster images from 142 countries paired with 33 million structured annotations spanning dense captioning with localization, incident assessment, relational scene descriptions, and crisis-grounded visual question answering (VQA) across diverse hazard types and scene contexts worldwide. Unlike prior disaster benchmarks that focus narrowly on classification, DisasterVLG unifies recognition, grounded reasoning, and uncertainty-aware interpretation. A distinguishing feature is its evidence-grounded annotation design: each VQA instance includes supporting evidence text that links visual observations to answers, alongside positive and negative question pairs, multiple answer formats, and difficulty labels. Visual grounding layer links caption phrases to image regions and relationship statements to subject-object boxes, with explicit labels for visual supports and clues. To support rigorous evaluation, we construct a human-verified test set of 38K images (1.15M annotations) spanning broad geographic regions and disaster types. Experiments with domain-adapted vision-language models demonstrate strong performance across disaster captioning, incident recognition, and grounded VQA, while also exposing the limitations of general-purpose models in crisis-specific reasoning. DisasterVLG establishes a scalable, reproducible benchmark for vision-language models that must reason reliably under real-world uncertainty. We will publicly release the dataset, trained models, and code.

cs.CV

新论文 · v1arXiv:2611.15459

Fear of Memorizing Objects: End-to-End Zero-Shot Model-Based Object Detection

Leon Greiner, Wout Boerdijk, Junwen Huang, Benjamin Busam, Rudolph Triebel, Maximilian Durner

原文摘要

This paper presents FOMO (Fear of Memorizing Objects), an end-to-end trainable transformer-based architecture for zero-shot model-based object detection. Existing approaches typically follow a proposal-and-match pipeline: target-agnostic mask proposals are generated and then matched to template features in a set of loosely coupled stages, making them susceptible to fragmented proposals in clutter and degraded matching under occlusions. FOMO elegantly replaces this paradigm by injecting object template information early in the detection pipeline: Bidirectional attention between learnable latent tokens, multi-view CAD templates, and query image features produces object-aware image features and conditioned queries which are subsequently decoded into bounding-box and confidence predictions. We train FOMO on Clutter6D, a novel large-scale synthetic dataset with over 500k images featuring more than 10k diverse object models rendered in highly cluttered scenes with multiple instances and varying occlusion levels, which addresses the limited object diversity and lack of realistic dense clutter in existing datasets for training zero-shot multi-instance algorithms. In true zero-shot fashion, FOMO achieves 54.2 mAP on BOP-Classic-Core at 0.23 s per image, ranking second overall with 3.8 pp difference to the top method while being more than 7x faster. Throughout comprehensive ablation studies we investigate and verify architectural design choices.

cs.CV

新论文 · v1arXiv:2611.15445

ReAlign: Reward-Guided Preference Alignment for Generative Re-Ranking

YaChen Yan, Liubo Li

原文摘要

Re-ranking refines the upstream ranked list to maximize listwise user utility. Existing two-stage generator-evaluator methods decouple list generation from list evaluation, leaving the generator without direct access to the listwise utility signal, while scoring-based one-stage methods score items under a fixed upstream permutation context without explicitly searching the permutation space. We propose \textbf{ReAlign}, a reward-guided preference alignment framework that couples a listwise reward model with a candidate-aware autoregressive generator. The reward model decomposes the listwise reward into position-wise contributions, modeling user exposure through a survival-based exposure module and item value through per-objective heads for heterogeneous engagement signals. The generator is trained in two stages: supervised pre-training on logged exposure lists to establish a relevance prior, followed by reward-guided post-training that aligns the generator with the reward model through complementary distribution and ranking alignment over within-pool standardized rewards. Extensive offline experiments show that ReAlign consistently outperforms existing state-of-the-art baselines.

cs.IR

版本更新 · v2arXiv:2611.12680

Wolfe's Minimum-Norm-Point Method Generalizes the Simplex Method

Yann Disser, Nils Mosis

原文摘要

Wolfe's method computes the point of minimum Euclidean norm in the convex hull of finitely many points. The method is parameterized by an insertion rule and it has been an open problem whether Wolfe's original rule guarantees a polynomial number of iterations (De Loera, Haddock, and Rademacher; STOC'18). We resolve this question by proving an exponential lower bound. In fact, we establish a tight connection to the simplex method for linear programming that allows us to transfer many lower bounds for simplex pivot rules to corresponding insertion rules of Wolfe's method. These include not only Wolfe's linopt rule via Dantzig's pivot rule, but also many other classical, randomized and history-based rules. As a corollary, we obtain the first exponential example for the Lawson-Hanson algorithm for nonnegative least squares, as well as PSPACE-hardness of deciding whether the linopt rule ever uses a given point.

math.OC · cs.DS