Audio and Speech Processing
Audio Preprocessing Effects on Stuttering Detection: A Class-Specific Analysis
Anisha Pattanayak, Hanie Kang, Huang-Cheng Chou et al. (+1 more)
Audio preprocessing can affect how well a system detects stuttering. We study a simulated chain of denoising,loudness normalisation, Opus coding, and voice activity detection on SEP-28k. We use frozen WavLM Base+ features and report pointwise confidence intervals from episode-level bootstrap resampling. At a fixed threshold of 0.5, the chain reduces block F1 from 0.638 to 0.465, with smaller decreases for the other four classes. ROC-AUC decreases for all five classes. Blocks show the largest F1 and ROC-AUC losses, while sound repetitions show the largest average precision loss. Tuning the threshold on processed validation audio raises block F1 to 0.630. Retraining on processed audio with threshold tuning gives 0.628. Threshold adjustment therefore accounts for most of the observed block F1 recovery. It does not change ROC-AUC, which retraining raises only from 0.620 to 0.633, compared with 0.724 on clean audio.
Read the full abstract ▾
All authors Anisha Pattanayak, Hanie Kang, Huang-Cheng Chou, Sudarsana Reddy Kadiri
Astrophysics of Galaxies
The Physical Origins of Scatter in the Dust Attenuation Scaling Relation of Star-Forming Galaxies: Star–Dust Geometry
Zongfei Lyu, Xian Zhong Zheng, Zhizheng Pan et al. (+11 more)
Characterising the physical drivers of
dust attenuation in
star-forming
galaxies (SFGs) is essential for interpreting their spectral energy distributions and
star formation histories. The infrared excess (IRX$\equiv$$L_{\rm IR}/L_{\rm UV}$) follows a universal scaling relation with
metallicity,
star formation rate,
galaxy size, and inclination. However, the origin of the scatter around this relation remains poorly understood. We revisit this relation using $\sim$32,000 local SFGs from SDSS, GALEX, and WISE, and investigate why some
galaxies deviate systematically from the best-fit relation. We find that the deviations are systematically linked to UV luminosity. UV-faint SFGs ($\log(L_{\rm UV}/{\rm L}_\odot)\le 9$) exhibit a median IRX excess of +0.23 dex, while UV-bright SFGs ($\log(L_{\rm UV}/{\rm L}_\odot)\ge 10$) show a median deficit of $-$0.20 dex relative to the relation defined by the dominant UV-intermediate population (86.38 percent of the sample). These offsets are not driven by
metallicity, total infrared luminosity, or specific
star formation rate. Instead, the deviations are closely linked to systematic differences in
star–
dust geometry. UV-faint SFGs are compact (median $R_{\rm e} = 2.90$ kpc) and preferentially viewed edge-on (median $b/a = 0.41$), leading to high effective
dust column densities along the line of sight. UV-bright SFGs are extended (median $R_{\rm e} = 5.99$ kpc) and preferentially viewed face-on (median $b/a = 0.74$), allowing UV
photons to escape efficiently.
Galaxies with significant bulge components ($B/T > 0.4$) populate the envelope of the relation, exhibiting larger scatter. We conclude that the scatter in the IRX relation can be largely explained by variations in the three-dimensional
star–
dust geometry, with UV luminosity acting as an effective tracer of these
geometric and physical differences.
Read the full abstract ▾
All authors Zongfei Lyu, Xian Zhong Zheng, Zhizheng Pan, Antonios Katsianis, Stijn Wuyts, Haiguang Xu, Pedro Cataldi, Wenhao Liu, Man Qiao, Dong Dong Shi, Yuheng Zhang, Shuang Liu, Run Wen, Chao Yang
Distributed Computing
AgentReplay: Token-Wise Trace Replay Is Essential for Fair Serving System Performance Benchmarking
Zaifeng Pan, Michael Wang, Chris Wu et al. (+4 more)
LLM-based
agents execute multi-turn workflows with interleaved model inference and tool calls, making efficient serving increasingly important. However, evaluating serving optimizations is challenging because identical tasks can produce different execution trajectories. Changes in generated
tokens can alter subsequent
prompts, tool calls, and
reasoning turns, making it difficult to distinguish system improvements from workload variation. Greedy decoding does not guarantee identical outputs, while replaying only sequence lengths loses
token information that affects prefix caching and mixture-of-experts (MoE) routing. To address these problems, we propose AgentReplay, a configurable trace record-and-replay framework for
agent serving. AgentReplay records input/output
tokens, expert selections, request dependencies, and tool durations. During replay, it forces the recorded output
tokens while performing normal autoregressive computation, with optional controls for MoE expert selection and tool delays. This allows different systems to execute the same recorded workload while retaining their own batching, scheduling, and parallelization decisions. We further separate trajectory generation from performance evaluation, enabling compatible smaller models to replay long-horizon traces collected with more capable models. Our experiments show that
token-wise replay in AgentReplay effectively eliminates workload variation that greedy decoding and length-wise replay cannot avoid, enabling fairer performance comparisons across serving configurations.
Read the full abstract ▾
All authors Zaifeng Pan, Michael Wang, Chris Wu, Zhengding Hu, Xinwei Qiang, Zhongkai Yu, Yufei Ding
Information Theory
Certified Bayesian optimal experimental design: from Expected Information Gain to Signal-to-Noise Ratio
Mohamed Doumbouya, Arthur Vidard and Olivier Zahm
Optimal experimental design is often formulated as the maximization of the Expected Information Gain (EIG), which measures the expected reduction in uncertainty about a parameter of interest after observing the data. For nonlinear forward models and nonGaussian priors, computing the EIG typically requires costly nested Monte Carlo
estimators or sophisticated density-approximation techniques. In this work, we derive new upper and lower bounds on the EIG using dimensional logarithmic Sobolev inequalities. The proposed bounds are expressed in terms of signal-to-noise ratios (SNR), preserving the simple
matrix-based structure of the EIG in the linear-Gaussian setting and depending solely on covariance
matrices and Fisher information
matrices. We introduce two families of bounds: conservative bounds, designed to retain the entire information content, and incremental bounds, that prioritize early information acquisition for sequential experimental design. When model gradients are available, we further derive computable approximations of the Fisher-information terms that require only a small number of gradient evaluations, yielding substantial computational savings. The tightness of these bounds is controlled by the nonlinearity of the forward model, and in the linear-Gaussian setting, these bounds recover the exact value of the EIG. By eliminating the nested-sampling bottleneck, the proposed framework offers a tractable and theoretically grounded alternative to direct EIG estimation for large-scale
Bayesian experimental design, while retaining the interpretability of the SNR-based bounds.
Read the full abstract ▾
Computation and Language
Supporting and Performing Culture from the Inside
Lea Frermann and Steven Bird
Anthropology and related disciplines which study culture have often found it useful to consider their epistemological and methodological approaches in terms of emic versus etic. This is the distinction between the insider perspective, how the world is viewed by a member of the culture, versus the outsider perspective, the scientific cataloguing of cultural knowledge and practice. We adopt this framework to systematically analyse assumptions in recent research on culture in NLP along the full pipeline of task selection, data collection, system design and evaluation. We draw
attention to the `culture of NLP' which shapes the research focus and approaches of our field, and suggest pathways towards a more culturally attuned AI.
Read the full abstract ▾
Computer Vision
HeroFrame-Bench: Reference-Anchored Evaluation via Rubric–Ranking Co-Evolution for Movie Hero Frame Selection
Weitai Kang, Hanieh Deilamsalehy, Yumo Xu et al. (+3 more)
Hero frames are in-film stills used as source imagery for theatrical posters, streaming cover art, film database listings, and other promotional placements. As the first visual entry point, they shape audiences' initial impressions of the movie and their subsequent willingness to watch it. Selecting these frames, a task we term hero frame selection, requires balancing content relevance with aesthetic appeal. A related task is keyframe selection, yet its
benchmarks prioritize relevance over aesthetics, using either finite annotations that exclude valid alternatives or VideoQA that entangles selection quality with downstream model capability. We therefore introduce HeroFrame-Bench, built through a scalable VLM-as-a-Judge framework. We construct
multimodal contexts from diverse metadata to ground a VLM judge that scores selected frames using our Reference-anchored Percentile. The percentile is obtained by inserting each frame into reusable, pre-ranked reference chains, enabling direct, extensible, and reliable evaluation. To reduce ambiguity and improve consistency in these subjective judgements, we further propose Rubric-Ranking Co-
Evolution, which generates movie-specific rubrics to condition the VLM judge and refines rubrics jointly with the resulting rankings. Within this process, we introduce several verifiable signals, most notably the Inverted Rubric Attack, to select robust rubrics. Finally, HeroFrame-Bench is instantiated over 204 movies with 2,031 reference chains and 1,970 learned rubrics. We build an annotation interface for human-
alignment studies which show that our construction design improves VLM agreement with human from 77.56% to 83.78%. Evaluation on multiple methods show that hero frame selection remains challenging.
Read the full abstract ▾
All authors Weitai Kang, Hanieh Deilamsalehy, Yumo Xu, Dewang Sultania, Serdar Cellat, Yan Yan
Machine Learning
SIMANF: Sample Free Learning of Unnormalized Distributions via Simulated Annealing in Normalizing Flows
Vikas Kanaujia
Efficiently learning and sampling from high dimensional,
multimodal unnormalized distributions without target samples remains a challenging problem. Although normalizing flows can generate samples efficiently, training based on the reverse KL divergence using only the unnormalized target density may suffer from mode collapse. We introduce SIMANF, a sample free framework that integrates simulated annealing with normalizing flows. SIMANF progressively transforms the target distribution from a smooth initial form to the original target distribution and trains the flow sequentially across these stages. By transferring the learned representation between stages, the method promotes mode coverage while progressively capturing finer features of the target distribution. Following annealing, a final refinement stage combines the reverse KL divergence with an importance weighted forward KL objective using samples generated by the flow. SIMANF requires no target samples during training and uses only the unnormalized density. We demonstrate its effectiveness on Many-Well distributions and high dimensional Scalar Phi4
lattice field theory distribution.
Read the full abstract ▾
Distributed Computing
RR-Evict: Fine-Grained Prefix Cache Eviction beyond LRU for Agentic LLM Serving
Zaifeng Pan, Chris Wu, Zhengding Hu et al. (+3 more)
LLM-based
agents execute long-horizon tasks through repeated model calls interleaved with tool execution and user interaction. As each call extends the history accumulated in previous turns, prefix caching avoids repeated prefill of the
agent's entire context. However, the aggregate cache footprint grows with context length and concurrency, forcing serving systems to reclaim cached KV tensors. We identify recency synchronization: accesses to an
agent's cached history refresh its cache nodes together, causing least recently used (LRU) eviction to concentrate on a few
agents' private histories. When a fully evicted
agent returns, it must recompute nearly its entire accumulated context, producing a large time-to-first-
token (TTFT) outlier even if most other requests retain substantial cache reuse. We present RR-EVICT, a fine-grained prefix-cache eviction strategy that distributes reclamation across idle
agent trajectories. RR-EVICT visits
agents in round-robin order and evicts a tail chunk from each, preserving reusable prefixes for more
agents when capacity remains for idle state. Returning
agents can reuse these partial histories and recompute smaller missing suffixes. The policy requires no prediction of future arrivals or tool latency. We implement RR-EVICT in SGLang and evaluate conversational and coding-
agent workloads under colocated and prefill-decode-disaggregated serving. Compared with LRU, RR-EVICT reduces P99 TTFT by up to 75.4% and P99 uncached
prompt tokens by up to 65.7%.
Read the full abstract ▾
All authors Zaifeng Pan, Chris Wu, Zhengding Hu, Xinwei Qiang, Zhongkai Yu, Yufei Ding
General Relativity and Quantum Cosmology
Stability Analysis of Canonical Acoustic Black Hole
HengZhong Fang, LiQiang Miao and ChangSen Liu
In this paper, we investigate the stability of canonical acoustic black hole under the perturbation of non-minimally coupling scalar field, electromagnetic field and Dirac field. The quasi-normal modes(QNMs) are computed using the three-order WKB approximation method. The results show that, for all these three kinds of perturbation, the values of the effective potential are positive, while the imaginary parts of QNMs are negative, the canonical acoustic black hole is a table. In details respectively, for scalar field perturbation, the effective potential and the QNMs depend on the angular
quantum number $l$, the overtone number $n$, and the coupling parameter {\zeta}, the perturbation decays more rapidly for bigger {\zeta}. For electromagnetic field perturbation, the effective potential and QNMs depend on $l$ and $n$. For Dirac field perturbation, the effective potential and QNMs depend on $n$ and the total angular
quantum number | k |.
Read the full abstract ▾
Machine Learning
HyperLabel: Multi-Label Classification via Hypergraph-Based Label Correlation Modeling
Peiyu Zhang, Heng Ping, Nikos Kanakaris et al. (+5 more)
Multi-label classification (MLC) requires predicting multiple relevant labels for each instance, where a central challenge is modeling complex label dependencies arising from co-occurrence patterns. Existing approaches are limited in capturing high-order label correlations, relying on implicit learning through contrastive objectives or pairwise
attention mechanisms without structural guidance. We propose HyperLabel, an encoder-decoder framework that explicitly models label dependencies through hypergraph neural networks. Our contributions are twofold: (i) We construct a label hypergraph where sample-defined hyperedges naturally encode multi-way co-occurrence patterns, providing explicit structural prior knowledge that captures relationships beyond pairwise interactions. (ii) We propose a unified cross-modal learning approach where HGNN+ performs bidirectional message passing to integrate feature information with label structure, and a shared cross-
attention decoder processes both modalities through complementary learning objectives. Extensive experiments on seven
benchmark datasets demonstrate that HyperLabel achieves state-of-the-art performance, with particularly significant improvements on macro-F1 scores (+10.3% on Delicious, +8.2% on Bibtex), validating that explicit hypergraph structure effectively captures complex label relationships. The code is available at https://github.com/iZHpy/Multi-label_hypergraph .
Read the full abstract ▾
All authors Peiyu Zhang, Heng Ping, Nikos Kanakaris, Yucheng Zhao, Shixuan Li, Wei Yang, Xiongye Xiao, Paul Bogdan
Machine Learning
Haifang Cao, Boan Tao, Xiyuan Gao et al. (+3 more)
Hyperbolic geometry has emerged as a principled space for representing hierarchical graphs. However, existing
hyperbolic graph neural networks typically rely on shared
curvature configurations and feature-driven
attention, failing to explicitly exploit local hierarchical topological patterns. To bridge this gap, we introduce the
Hyperbolic Sombor Index (HSO) as a lightweight structural prior for capturing hierarchy-indicative degree stratification. Building on this, we propose HSO-GAT, a
topology-adaptive
hyperbolic graph
attention network that unifies
geometric adaptation and message propagation. Specifically, it comprises two complementary modules: HSO-Guided Local
Curvature Adaptation, which performs adaptive node-wise
geometric scaling from aggregated node-level HSO signals, and HSO-Gated
Hyperbolic Graph
Attention, which enables structure-aware message passing through feature-conditioned gating. Theoretically, we establish the monotonic sensitivity of edge-level HSO to degree imbalance and analyze the validity and radial scaling properties of node-adaptive
hyperbolic mappings. Extensive experiments on eight
benchmark datasets demonstrate that HSO-GAT consistently achieves state-of-the-art performance in both node classification and link prediction tasks.
Read the full abstract ▾
All authors Haifang Cao, Boan Tao, Xiyuan Gao, Timing Li, Yu Wang, Pengfei Zhu
Artificial Intelligence
When Does a Skill Add Value? Task-Conditional Gain Prediction for Selective Skill Use
Anjie Xu, Zhiyu Zhang, Ruiqing Ding et al. (+2 more)
Agent skills are expected to improve task performance. Yet we find that they often provide no benefit, and can even hurt performance while incurring additional
token costs. Can we predict whether a skill will help before the
agent acts? We introduce SkillDelta, a framework for predicting task-conditional skill gains from paired executions of the same
agent with and without the skill. A local predictor transfers these historical gains to new tasks without retraining the
agent. Under explicit transfer assumptions, our analysis links support coverage, representation mismatch, and execution noise to prediction error and decision regret. Across five
benchmarks and three target
agents, paired history improves observed-gain ranking over skill-assisted outcomes alone in 12 of 15 settings. At matched expected skill-use rates, SkillDelta improves success over random activation in all 15 settings, with an average absolute gain of 4.3%. Most of this advantage comes from allocation across task groups. Evidence for additional within-group selection value is strongest on ToolQA and weaker elsewhere. Code is available at https://github.com/TankTechnology/skilldelta.
Read the full abstract ▾
All authors Anjie Xu, Zhiyu Zhang, Ruiqing Ding, Fengli Xu, Leye Wang
Computer Vision
FSS-UBrain: Multi-region Few-Shot Brain Tumor MRI Segmentation
Truong Viet Vu, Nguyen Phuc Nguyen, Dang Thi Thu Hang et al. (+4 more)
Accurate delineation of whole tumor (WT), tumor core (TC), and enhancing tumor (ET) from
multimodal magnetic resonance imaging remains challenging under limited annotation, cross-cohort variation, and severe target sparsity. We propose FSS-UBrain, a region-wise one-shot
segmentation framework that uses a labeled positive support slice to condition binary query
segmentation separately for WT, TC, and ET. Support-derived foreground and background descriptors guide query-feature adaptation, bottleneck interaction, decoder-side reconstruction, and boundary refinement. Episodic training additionally incorporates hard-negative and fully negative queries with empty-query regularization to suppress spurious foreground activation when the selected region is absent. Although inference operates on two-dimensional support–query slice pairs, checkpoint selection, threshold calibration, and final evaluation are performed after volumetric reconstruction. FSS-UBrain is evaluated on a held-out BraTS 2020 split and under target-supported cross-cohort protocols on BraTS 2023 and BraTS-Africa. Cases used as target support are excluded from the query cohorts, and no target-domain fine-tuning or test-time parameter updates are performed. On BraTS 2020, FSS-UBrain achieves volumetric Dice scores of 89.82%, 82.14%, and 77.42% for WT, TC, and ET, respectively, with corresponding 95th-percentile Hausdorff distance (HD95) values of 11.12, 9.01, and 4.46 mm. It also achieves the highest mean Dice and lowest finite-pair mean HD95 point estimates on BraTS 2023 and BraTS-Africa among the compared few-shot methods. These findings support target-conditioned few-shot
segmentation while highlighting sensitivity to support selection and cohort-specific variation.
Read the full abstract ▾
All authors Truong Viet Vu, Nguyen Phuc Nguyen, Dang Thi Thu Hang, Tran Thien Thanh, Vo Nguyen Quoc Bao, Nguyen Thai Anh, Ngo Hoang Tu
Machine Learning
Learning Through Game: Skewed Transfer of Tabular Knowledge to Strengthen Image Model
Longfei Huang, Shangdong Yang and Yang Yang
Multimodal tabular-image learning is gaining growing
attention, yet it faces challenges due to tabular data unavailable at test time. A practical solution involves transferring tabular knowledge to images during training to enhance the performance of image models at inference. However, the overlooked yet important challenges lie in the modality imbalance between images and tables, as well as their asymmetric modality relationship in cross-modal transfer, which limits the auxiliary role of tabular data. To address these issues, we propose Skewed Knowledge Transfer (SKT), which asymmetrically transfers tabular knowledge to improve the image model by adaptive integration of modality gradients in a shared parameter space. Specifically, we first introduce a
multimodal shared head, which allows the model to benefit from cross-modal structure without adding additional parameters. We then design a two-step Nash Bargaining strategy to effectively leverage tabular gradients. In the first step, SKT seeks a point of modality balance and uses preference awareness in the second step to steer combined gradients toward image-beneficial directions. Furthermore, we theoretically analyze the Pareto improvement and convergence of SKT. To this end, tabular knowledge is explicitly transferred to enhance image models. Empirical experiments on widely used tabular-image
datasets reveal that SKT consistently improves image unimodal performance by using tabular data as auxiliary information.
Read the full abstract ▾
Machine Learning
Certification Frontiers for Gaussian LoRA: Independent Priors, Posterior Risk, and Prediction-Preserving Balancing
Joyanta Jyoti Mondal and Ibne Farabi Shihab
Post-hoc
Bayesian fine-tuning places Gaussians around trained low-rank adapters, yet a calibrated posterior does not by itself yield a useful generalization certificate. Such a posterior admits an informative PAC-Bayes certificate only when both the loss of its sampled predictors and its KL divergence from an admissible prior are small. In this research, we characterize this certification frontier for Gaussian LoRA posteriors and separate three interventions: changing the prior, changing the
stochastic predictor, and changing only how its complexity is counted. First, an exact isotropic KL envelope eliminates the prior scale and yields a width threshold that excludes posterior widths before sampling, while a zero-KL floor identifies targets that no complexity reduction can reach at a measured risk bound. Second, we minimize KL in closed form over the full $GL(r)$
symmetry of the low-rank factors, leaving every sampled adapter product unchanged, and derive the noncentral objective required when the prior center is trained on an independent split. On a small-pool RoBERTa audit of 567 configurations, the recorded 64-draw summaries imply a certificate floor of $0.7298$ even with zero KL and Chernoff accounting, so reducing complexity alone cannot certify these posteriors at the recorded budgets. For a stable posterior in a controlled Gaussian-factor task, Chernoff accounting certifies risk below $0.1$ on 20 of 20
datasets with 1024 draws, whereas Hoeffding certifies none. On deliberately deformed synthetic rank-four factors,
matrix balancing reduces KL by $29.3\%$ on average beyond scalar balancing without changing any prediction. Numerical split-prior scenarios make the remaining risk and complexity budgets explicit.
Read the full abstract ▾
Machine Learning (Statistics)
Overfitting of Spectral Gradient Descent: How Matrix Geometry shapes Generalization and Implicit Bias
Guillaume Braun, Ichiro Hashimoto and Masaaki Imaizumi
We study the generalization of spectral gradient descent (SpecGD) in overparameterized
matrix classification with corrupted labels. Each input combines a shared low-rank signal with a rank-one sample-specific perturbation, referred to as a shortcut, that enables memorization but does not generalize. We contrast collapsed shortcuts, which share a singular direction, with dispersed shortcuts, which occupy distinct singular directions. Changing only this geometry can reverse the relative generalization of GD and SpecGD: collapsed shortcuts can favor SpecGD, while dispersed shortcuts can favor GD. In the dispersed regime, exact shortcut orthogonality eliminates the signal from the late-stage SpecGD direction, while vanishing random correlations collectively generate a small but generalization-relevant signal through a second-order effect. To identify the direction selected by SpecGD, which the spectral max-margin problem alone does not determine, we combine a refined analysis of its dual with the exponentiated-gradient dynamics of normalized loss weights. Finally, we show that a single SpecGD step can already interpolate and generalize well, while continued training converges to a direction with substantially worse generalization.
Read the full abstract ▾
Computation and Language
Before Answering: Evidence Sufficiency under Size-Matched Memory Construction
Joyanta Jyoti Mondal, Md. Shifatul Ahsan Apurba, Mridul Banik et al. (+2 more)
Agents that answer questions from compressed or retrieved
memory must recognize when the evidence a query needs is no longer in
memory.
Benchmarks for this task usually create insufficient-evidence examples by deleting supporting passages. We show that this construction leaks the label through
memory size: on MuSiQue, a classifier that only counts paragraphs reaches an area under the ROC curve (AUROC) of $0.979$ for detecting unsafe
memory, higher than the lexical
estimator we initially evaluated. We propose a size-matched construction that provably removes this shortcut, and use it to study MemSafe, an
estimator that cross-encodes the query with each
memory unit and aggregates the units with a set
transformer. Across three multi-hop question answering
datasets and five seeds, MemSafe reaches $0.968$ and $0.983$ AUROC on MuSiQue and HotpotQA, $0.26$ to $0.39$ above a lexical baseline, while the third
dataset, 2WikiMultiHopQA, is saturated. A frozen pretrained cross-encoder with a logistic head already closes $41\%$ of the MuSiQue gap between the lexical baseline and MemSafe. At the same time, MemSafe degrades more than a weak baseline on the unanswerable questions released with MuSiQue, reaches only $0.639$ AUROC on SQuAD~2.0, and needs several thousand clinical training examples before it outperforms a feature-based
estimator. Used as a gate for a 7B reader, it reduces the error rate on answered questions from $0.850$ to $0.631$ at $5\%$ coverage, outperforming both reader confidence and, on average, the ground-truth integrity label, although a 7B LLM judge is the better gate at $10\%$ coverage. These results indicate that the way insufficient evidence is constructed matters as much as the
estimator that detects it.
Read the full abstract ▾
All authors Joyanta Jyoti Mondal, Md. Shifatul Ahsan Apurba, Mridul Banik, Md Masud Al Mahmud, Ibne Farabi Shihab
Machine Learning
Self-Reconstruction Dynamics for Autoencoder Reconstruction Refinement
Hitoshi Iyatomi
Standard autoencoder (AE) inference uses a single encoder-decoder pass, though the latent may not be optimal for each sample under a fixed decoder. We ask whether a trained AE can reveal information for improving its own reconstruction. Repeated application of a frozen AE to its reconstruction produces transient image- and latent-space trajectories, termed Self-Reconstruction Dynamics (SRD). Although this degrades fidelity in the AEs studied here, SRD contains sample-specific information for correcting the reconstruction. We propose SRD-guided Reconstruction Refinement (SRD-RR), which predicts a latent correction from a short SRD with the AE frozen and no per-sample test-time optimization. We also introduce MSE-recov, an MSE recovery ratio relative to an empirical decoder-optimized reference. Across six
datasets, SRD-RR recovers 38.6% of the empirically recoverable MSE gap with one transition and 45.3% with two. A two-transition variant trained without direct access to original images, using an SRD-derived pseudo-target, achieves 40.7% recovery and a 1.74 dB average PSNR gain. Removing trajectory information reduces the gain, while cross-sample trajectory assignment causes severe degradation, confirming strong sample specificity. Nonlinear SRD-conditioned refinement consistently outperforms fixed and trained linear latent correction. On a pretrained DINOv2-based representation autoencoder (RAE) with substantially different latent dynamics, SRD conditioning again improves a matched trajectory-free predictor. However, pixel-MSE latent refinement reveals a strong mismatch between pixel fidelity and perceptual quality, while the SRD-derived pseudo-target mitigates this degradation. Overall, SRD is a useful sample-specific refinement signal, while the objective determines how it translates into pixel and perceptual quality.
Read the full abstract ▾
Machine Learning
When Can Old Evaluations Certify a New Model? Label-Efficient Release Decisions under Evaluator Drift
Joyanta Jyoti Mondal, Mridul Banik, Md. Shifatul Ahsan Apurba et al. (+1 more)
Releasing a model update requires certifying that its current-population risk stays below a threshold. Trusted labels are expensive, while a cheap evaluator, such as an LLM judge, scores every example. Reusing evaluator errors from earlier audits is tempting, but when may such evidence replace current labels? It depends on the status of history. If the errors can change invisibly, no label-free test detects the change, and every valid, useful certifier must keep buying labels at a rate we characterize; if a bound on the change is assumed, label-free certification is valid at an explicit error cost. For the middle ground, where history is informative but untrusted, we propose portfolio vigilance, a sequential certifier mixing a betting expert guided by history with one that learns only from current labels; history affects only how it bets, so validity holds for any history. The contribution is not prior-informed betting or expert mixtures, but separating history that may enter validity from history that may only guide label collection. In a canonical model, accurate history shortens decisions but never raises the evidence growth rate; stale history can destroy it. On held-out CIFAR-10N and DICES-990 data, portfolio vigilance needs 0.465 (95% CI $[0.327,0.575]$) and 0.740 ($[0.618,0.877]$) times the labels of a matched prediction-powered monitor, with no observed false certification, and fewer labels on all six external blocks. Under corrupted advice it stays within 8.0% of its better component, while trusting history alone costs up to 1.66 times as much. In post-confirmatory repeated-judge experiments on DICES-990 and ToxicChat, changing a fixed LLM judge's rubric moves its scores beyond run-to-run variation; the portfolio then needs 0.790 ($[0.667,0.909]$) and 0.631 ($[0.520,0.770]$) times the labels of the matched monitor, and fewer than trusting history alone.
Read the full abstract ▾
All authors Joyanta Jyoti Mondal, Mridul Banik, Md. Shifatul Ahsan Apurba, Md Masud Al Mahmud
Strongly Correlated Electrons
Symmetry-permuting entanglers for non-invertible symmetry-protected topological phases
Minyoung You
1+1D
symmetry-protected topological (SPT) phases with invertible
symmetry can be characterized by the symmetric entanglers: globally symmetric finite-depth local
unitary circuits that generate representative ground states from a product state. For non-invertible
symmetries, however, distinct SPT phases, such as the three $\mathrm{Rep}(D_8)$ SPT phases, may not be connected by any symmetric entangler. We show that the three $\mathrm{Rep}(D_8)$ SPT phases can nevertheless be connected by
symmetry-permuting finite-depth local
unitary circuits, realizing dualities predicted by topological
quantum field theory (TQFT). We construct an explicit
matrix product
unitary connecting two of the $\mathrm{Rep}(D_8)$ SPT phases, and show that its action on string order parameters reproduces the anyon permutation predicted by the TQFT duality.
Read the full abstract ▾
Plasma Physics
Review of LHD-ECRH activities and beyond
Masaki Nishiura, Hiroe Igami, Hiromi Takahashi et al. (+11 more)
The electron cyclotron resonance heating (ECRH) system on the Large Helical Device (LHD) played a key role in the progress of helical fusion research from the first
plasma in 1998 to the final discharge on December 25, 2025. This paper presents a historical review of ECRH activities on LHD, highlighting the stepwise advances in hardware, including gyrotrons, transmission lines, and launcher systems, together with the exploration of advanced
plasma regimes such as overdense heating and long-pulse steady-state operation. Beyond
plasma heating, the ECRH system was also used as a high-power probe for advanced diagnostics, including collective Thomson scattering (CTS) as well as correlation electron cyclotron emission (CECE), and as a flexible actuator for transport studies and real-time control experiments. Finally, we discuss how the technological assets and operational expertise accumulated over the nearly three decades of the LHD program are being further developed and applied to next-generation helical devices, particularly CHD and CHD-U, as well as to broader gyrotron and millimeter-wave applications.
Read the full abstract ▾
All authors Masaki Nishiura, Hiroe Igami, Hiromi Takahashi, Naoki Kenmochi, Ryoma Yanai, Kenji Ueda, Yoshinori Mizuno, Toshiki Takeuchi, Yasuo Yoshimura, Takashi Shimozuma, Shin Kubo, Toru Ii Tsujimura, Sakuji Kobayashi, Tsuyoshi Kariya
Computation and Language
LANTERN: Illuminating Hidden Mathematical Knowledge in Language Models
Pavel Tikhonov, Elena Tutubalina, Ivan Oseledets et al. (+2 more)
Language models can now prove
theorems, but people still decide which problems to pursue. We ask whether a model's internal representations can help identify promising mathematical connections. We develop LANTERN, a fast, cost-efficient pipeline that uses a classifier over pretrained-model activations to rank candidate relations, followed by staged filtering, hypothesis generation, executable verification, and analytical checking. Applied to the On-Line Encyclopedia of Integer Sequences (OEIS), LANTERN ranked 50 million pairs among 10,000 frequently referenced sequences and produced 62 verified relations between pairs without an existing OEIS cross-reference. A content screen retained 13 relations worth presenting; nine of these are informative or insightful, including four which are entirely novel to the best of our knowledge: none appears in the OEIS or in our targeted literature search. The entire end-to-end process including classifier training, candidate ranking, filtering and verification took under 8 hours.
Read the full abstract ▾
All authors Pavel Tikhonov, Elena Tutubalina, Ivan Oseledets, Dmitry I. Ignatov, Mikhail Seleznyov
Machine Learning
Representation Editing for Multimodal Test-Time Adaptation
Longfei Huang, Xiangyu Wu and Yang Yang
Multimodal test-time adaptation (TTA) aims to adapt a pretrained
multimodal model online to distribution shift across modalities using unlabeled test data, showing broad potential in real-world applications. However, existing methods primarily focus on adjusting fused features to bridge the source-target gap, lacking explicit control over intermediate representation misalignment, which is a key driver of performance drop under distribution shift. In this work, we tackle this challenge from the perspective of representation engineering. Unlike previous TTA methods that update fusion weights in place, we propose
FourIer Representation Editor (FIRE), a novel
multimodal TTA approach that directly edits semantically rich intermediate representations. Specifically, we first adopt representation editors into each intermediate layer of the unimodal encoders, enabling layer-wise calibration of unimodal representations. To further enhance the diversity and stability of the low-rank editing subspaces, each representation editor performs frequency domain mixing via the fast
Fourier transform to construct structured bases. Moreover, we introduce multi-level adaptation objectives to optimize these editors, jointly promoting cross-modal semantic
alignment, source-target statistical
alignment, and asymmetric prediction consistency. In this way, FIRE yields aligned unimodal representations for fusion and further improves prediction reliability. Extensive experiments on two widely used
multimodal benchmarks under various corruption types demonstrate the superiority of FIRE over existing
multimodal TTA methods.
Read the full abstract ▾
Quantum Physics
Contour-integral and Fourier transform based multivariable quantum eigenvalue transformation for commuting matrices
Shan Jiang and Dong An
We study the problem of implementing multivariable
matrix-valued functions on a
quantum computer and propose two
quantum algorithms for multivariable
matrix eigenvalue transformations acting on tuples of pairwise commuting
matrices. The first
algorithm is based on multivariable contour integrals and applies to arbitrary holomorphic functions on $\mathbb{C}^n$. The second
algorithm is based on high-dimensional
Fourier transforms and designed for smooth functions of commuting Hermitian
matrices. For both
algorithms, we discuss and analyze the complexity of their
quantum implementation leveraging
quantum singular value transformation, compression gadgets, and linear combination of
unitaries. We additionally study their hybrid
quantum–classical variants that reduce the number of required ancilla
qubits to logarithmic dependence on the number of variables, at the cost of increased query complexity. As an application, we show that multivariable
matrix polynomials can be implemented with no additional explicit degree dependence, depending only on global properties of the
polynomials.
Read the full abstract ▾
Materials Science
Dopant-modulated lattice softening drives drastic thermal conductivity reduction in \beta-FeSi2 thermoelectrics
Cuiping Zhang, Qingyong Ren, Yangfan Cui et al. (+15 more)
Suppressing
lattice thermal conductivity (K_lat) is pivotal for thermoelectric efficiency. While traditional strategies rely heavily on
phonon scattering from mass- and size-mismatches, we demonstrate a robust K_lat suppression mechanism driven by dopant-induced
lattice stiffness modulation. Through a comparative analysis of p-type (Mn) and n-type (Co, Ir) doping in the \beta-FeSi2 model system, we show that Co and Ir doping significantly reduce K_lat. Notably, Co doping achieves a ~71% reduction at 300 K even without significant mass and size contrast. By correlating transport data with
neutron powder diffraction, heat capacity, and Raman spectroscopy, we reveal anomalous
lattice expansion, a substantial reduction in Debye temperature, and marked vibrational
redshift and broadening. These systematic changes provide strong evidence for atomic-scale
lattice softening and a fundamental weakening of interatomic force constants, which synergistically lower
phonon group velocities and amplify anharmonic scattering. Our findings establish
lattice stiffness manipulation as a powerful strategy for thermal management, offering a distinct design pathway beyond traditional mass- and strain-fluctuation models.
Read the full abstract ▾
All authors Cuiping Zhang, Qingyong Ren, Yangfan Cui, Chen Chen, Songbai Hu, Shengnan Dai, Chin-Wei Wang, Wanju Luo, Dexiang Gao, Bao Yuan, Junying Shen, Fan Chen, Wei Xu, Yuting Li, Mingfang Shu, Xiaoli Huang, Pengfei Qiu, Jie Ma
Data Structures and Algorithms
Tree Search With Distributional Predictions
Michael Dinitz and Bob Dong
Learning-augmented
algorithms use machine-learned predictions to improve classical algorithmic guarantees when the predictions are accurate, while retaining rigorous performance guarantees when they are not. We study this paradigm for search on trees. Given a tree $T$ containing an unknown target vertex $t$, an
algorithm may query any vertex $v$ and learn which neighbor of $v$ lies on the unique path from $v$ to $t$. The goal is to find $t$ using as few queries as possible. We consider the distributional setting, in which the target is drawn from an unknown distribution $p$ and the
algorithm is given a predicted distribution $\widehat p$ of unknown quality. We give an
algorithm with expected query complexity $O\left(H(p)+k\log \eta \right)$, where $H(p)$ is the Shannon
entropy of the true distribution and $\eta$ is the earth mover's distance between $p$ and $\widehat p$ in the tree metric. We also provide a matching lower bound that shows our
algorithm is asymptotically tight. Finally, experiments on real-world and synthetic trees show that our prediction-based
algorithm can use substantially fewer queries than a simple baseline that trusts the prediction completely.
Read the full abstract ▾
Artificial Intelligence
Prefill-Free Cross-Family KV Cache Transfer for Heterogeneous Multi-Agent LLMs
Vincent-Daniel Yun, Woosang Lim, Haneul Yoo et al. (+3 more)
Recent multi-
agent LLM systems increasingly combine heterogeneous models for specialized
agent roles. However, text-based communication requires each receiver to prefill shared context already processed by the sender. Reusing the sender's key-value (KV) cache avoids this redundancy, but prefill-free transfer across model families must handle differences in tokenization, model depth, and KV representations. To address these issues, we propose HeteroFold, a prefill-free cross-family KV cache transfer method that keeps both the sender and receiver frozen. HeteroFold aligns model structures, maps the sender cache into the receiver space, and calibrates it to preserve receiver behavior. Across six transfer directions, HeteroFold achieves the best cache-transfer performance on all four long-context
benchmarks and most short-context settings. It also matches text-based communication on the multi-
agent benchmark. At 32K context length, Llama-3.1-8B$\rightarrow$Ministral-3-14B transfer is $10.7\times$ faster than Native Prefill and $1.18$–$1.47\times$ faster than the state-of-the-art prefill-free baselines, Dense Latent and KV Ridge. These results show that HeteroFold enables efficient cross-family KV reuse without receiver prefill.
Read the full abstract ▾
All authors Vincent-Daniel Yun, Woosang Lim, Haneul Yoo, Sungjoo Yoo, Murali Annavaram, Sai Praneeth Karimireddy
Information Theory
Discrete Diffusion for Port Selection in Multiuser Pinching-Antenna Systems
Wenxuan Sun, Mingjie Shao, Yanqing Xu et al. (+1 more)
Pinching-
antenna systems (PASS) reconfigure wireless channels by activating radiating elements at selected locations along dielectric waveguides. Port selection is a core issue in improving the performance of multiuser PASS systems. In this paper, we study joint port selection and precoder design for multiuser sum rate maximization, which leads to a large-scale nonconvex mixed-integer problem coupling discrete port selection with continuous precoder design. We propose a discrete
diffusion method that does not require optimal or near-optimal port selection solutions as training labels and recasts the resulting optimization problem as sampling from a target distribution. Instead, we learn a potential function from local objective differences between neighboring feasible realizations. The learned potential is incorporated into a Metropolis–Hastings sampling rule to guide a parallel
diffusion process while preserving the port selection constraints. Simulation results show that the proposed method approaches exhaustive search within a $0.30\%$ sum rate gap while achieving over $150\times$ speedup, and outperforms state-of-the-art methods including greedy and beam search. In particular, it also achieves an $18.8\times$ speedup over greedy search.
Read the full abstract ▾
All authors Wenxuan Sun, Mingjie Shao, Yanqing Xu, Ya-Feng Liu
Combinatorics
Non-commutative frieze patterns over quaternion algebras and other normed division rings
Michael Cuntz, Thorsten Holm and Peter Jorgensen
Non-commutative friezes have been introduced by Berenstein and Retakh and studied further by the authors. In this paper we consider non-commutative friezes over normed division rings, like for instance Hamilton's quaternions or more general quaternion
algebras. We address the fundamental question in the theory of friezes of whether over a certain subset there are finitely or infinitely many non-commutative friezes (with 1's on the boundary) for any height. As an application of a
theorem bounding the norm of quiddity entries we deduce that for every norm-finite subset of a normed division ring there are only finitely many such non-commutative friezes for every height. In particular this result applies to the Lipschitz quaternions and the Hurwitz quaternions of Hamilton's quaternions. We then study more generally non-commutative friezes over Lipschitz subrings of non-split quaternion
algebras $(a,b)_{\mathbb{Q}}$. We determine the frieze subrings for all $a,b<0$, and as a consequence we see that all such non-commutative friezes are known if $a\le -4$ and $b\le -4$.
Read the full abstract ▾
Artificial Intelligence
LAM: Efficient Lossy Agent Memory Framework With A Retrieval-Score Error Bound
Baixi Sun, Le Chen, Anjir Ahmed Chowdhury et al. (+10 more)
Agent memory grows as
agents read inputs, reason, and call tools. Longer histories increase inference cost and eventually exceed the context window. LLM-based summarization reduces this history but adds latency and provides no explicit bound on information loss. We propose LAM, a Lossy
Agent Memory system with three components: a deterministic deduplication rule with a substitution bound on
retrieval scores - a bound on score perturbation, not a certificate of unchanged ranking; a
memory manager that preserves the cached prefix and overlaps compaction with inference; and a performance model that estimates compaction costs before deployment. On 600
agent trajectories, LAM removes 22.47% of observation
tokens while retaining 99.984% of the measured gold-patch evidence. At a fixed deletion set, the performance model predicts a 71.4x-91.6x end-to-end speedup from removing records before prefill instead of deleting them from a prefilled context. That benefit comes from the schedule rather than the rule and applies to any prefix-preserving test.
Read the full abstract ▾
All authors Baixi Sun, Le Chen, Anjir Ahmed Chowdhury, Xiaolong Ma, Chih-Hsuan Yang, Mingze Xia, Syed Zawad, Sheng Di, Rajkumar Kettimuthu, Huihuo Zheng, Rajeev Thakur, Venkatram Vishwanath, Feng Yan