Machine Learning
Uncertainty Evaluation: Bridging Machine Learning and Measurement Science
Dario Petri and Luca Mari
Reliable uncertainty evaluation is essential for assessing the trustworthiness of machine learning (ML) predictions, particularly in safety-critical applications. This paper explores conceptual analogies between uncertainty evaluation in ML and measurement science, drawing on the
evolution of metrology from error-based approaches to the framework established by the Guide to the Expression of Uncertainty in Measurement (GUM). Particular
attention is devoted to the distinction between process-level and result-level quality and to the limitations of the conventional classification of ML uncertainty into aleatoric and epistemic components. It is argued that this dichotomy, similarly to the traditional distinction between random and systematic measurement errors, is inherently model-dependent and provides an insufficient basis for a unified uncertainty evaluation framework. Further analogies between machine learning and measurement field of knowledge are identified concerning distribution shifts, anomalous operating conditions, and the validity of uncertainty evaluation. Building on the proposed considerations, the paper advocates a metrology-inspired approach based on the systematic identification, quantification, and propagation of uncertainty contributions. By extending the methodology already established in measurement science to ML, this approach could offer a potentially rigorous, consistent, and transferable framework for evaluating the trustworthiness of individual predictions.
Read the full abstract ▾
Information Theory · Cross-listed from math.CO
Non-additive symmetric rank-metric codes
Marta Messia and Ferdinando Zullo
We study non-additive rank-metric codes contained in the space of symmetric
matrices over a finite field. Our main focus is on symmetric $2$-codes. For odd $q$, we introduce a quotient construction for $\mathrm{Sym}_3(q)$ in which rank-one differences are described by isotropic vectors of a nondegenerate quadratic form on $\mathbb F_q^4$. This leads to a correspondence with partial ovoids of the elliptic quadric $Q^-(5,q)$. Using a construction of Cossidente and Pavese, we obtain a non-additive symmetric $2$-code in $\mathrm{Sym}_3(q)$ of cardinality $q^4+2q^3$, improving the previously known general lower bound. We also give an explicit
matrix description of this family. Lastly, we present a block-based construction for symmetric $2$-codes that extends the three-dimensional improvement to all odd dimensions. In particular, for every odd $n\geq3$ and every
prime power $q>2$, we obtain \begin{equation*} A_q^sym(n,2) \ge q^{(n^2-1)/2}+q^{(n^2-3)/2}+q^{(n^2-9)/2}, \end{equation*} which is strictly larger than the maximum cardinality of an additive symmetric $2$-code with the same parameters. This yields the first infinite family (as the
matrix dimension $n$ varies) of non-additive symmetric $2$-codes whose size is larger than the largest achievable size of an additive code with the same parameters; before this, such a gap had only been established for $n=3$.
Read the full abstract ▾
Computation and Language
A Benchmark for Bangla Task-Oriented Dialogue Understanding
Afrina Alam Lamia, Sumiaya Sharif, Prottoy Modak et al. (+2 more)
Task-oriented dialogue understanding remains underexplored for Bangla, particularly in multilingual settings involving English and romanized Bangla (Banglish). We introduce BanglaTOD, a parallel tri-lingual
benchmark comprising 209 multi-turn dialogues across nine Bangladesh-specific service domains and 64 intents. Each conversational turn is aligned across Bangla, English, and Banglish, yielding 4,689
language-specific turn records annotated with intents,
token-level slots, state updates, and accumulated dialogue states. We establish context-aware baselines for joint intent classification and slot filling using mBERT, BERT-base-uncased, and BanglaBERT. Cross-
language analysis reveals substantial performance differences across models and linguistic representations, while removing dialogue history reduces intent accuracy by 19.0–25.7 percentage points. BanglaTOD provides a foundation for multilingual dialogue understanding and future dialogue state tracking research in Bangladesh-specific conversational applications. The
dataset is publicly available at \href{https://huggingface.co/
datasets/afri7here/BanglaTOD}{Hugging Face}.
Read the full abstract ▾
All authors Afrina Alam Lamia, Sumiaya Sharif, Prottoy Modak, Ohiduzzaman Rayhan, Nafis Sadeq
Emerging Technologies
Two-State Dynamical Synapses as a Physical Substrate for Equilibrium Propagation
Dongshu Liu and Julie Grollier
Equilibrium Propagation (EqProp) updates each synapse by comparing a local term at the free and nudged
equilibria. Because these two values occur at different times, a physical implementation requires temporary local
memory. We show that this
memory can emerge from a two-timescale synaptic model in which neural and volatile synaptic dynamics are jointly derived from a single energy function. A volatile state $w$ relaxes toward the product of the pre- and postsynaptic activities, while a non-volatile state $M$ stores the learned parameter. At
equilibrium, $w$ equals the local term entering the EqProp update. Using its net change between phases implements phase-based EqProp, while transferring its successive changes during nudging yields continual EqProp, in which the temporal comparison is carried by the synaptic dynamics themselves. On MNIST with a single-hidden-layer network, the phase-based model reaches $2.01 \pm 0.09\%$ test error versus $2.00 \pm 0.13\%$ for standard EqProp, and the continual model reaches $2.24 \pm 0.11\%$ versus $2.28 \pm 0.16\%$ for continual EqProp. The framework further specifies the dynamical requirements that a volatile–non-volatile synaptic device must satisfy to implement this learning mechanism.
Read the full abstract ▾
Machine Learning
From Policy Evaluation to Off-Policy Control: Provable Emergence of In-Context Q-Learning
Yingjun Mo, Yajie Zhao and Fang Kong
Transformers can implement reinforcement-learning
algorithms in context, yet how off-policy control emerges within
Transformers remains poorly understood. We study Q-learning, a foundational model-free control
algorithm that combines off-policy bootstrapping with greedy policy improvement. We prove that a two-layer
attention-only
Transformer can implement a finite-temperature form of Q-learning, where the temperature $\tau>0$ controls the approximation to greedy action selection, recovering window Q-learning as $\tau\to0$. We train the structured
Transformer on randomly generated MDP tasks using output-only supervision. Experiments show that the theoretically predicted parameter structure emerges during training, including the internal greedy action-selection mechanism, and that the trained
Transformer achieves performance comparable to Q-learning on unseen tasks. Our theoretical analysis further shows that the equivalent realizations form a solution
manifold and that, under suitable task-diversity conditions, the training iterates converge locally to this
manifold at an exponential rate. For the random-task model analyzed here, the local worst-case number of gradient steps needed to reduce the distance to the
manifold by a factor $\epsilon$ is $\Theta\left(\tau^{-5}\log(1/\varepsilon)\right)$, quantifying the increasing optimization cost as the learned update approaches window Q-learning.
Read the full abstract ▾
Programming Languages
OpenMQPI: An MLIR-based Pipeline to Offload Quantum Circuits via OpenMP
Ercüment Kaya, Ekin Devim Şahinkaya, Minh Chung et al. (+1 more)
Quantum computing introduces specialized software stacks and programming interfaces that are often tightly coupled to specific ecosystems and hardware backends, limiting interoperability and software reuse. For HPCQC application developers, this complicates the integration of
quantum accelerators into existing applications. Existing
quantum frameworks also use ecosystem-specific
circuit representations and compilation pipelines, hindering integration with established heterogeneous compiler infrastructures. We propose OpenMQPI, an MLIR-based OpenMP extension for
quantum programming that leverages OpenMP's standardized target offloading infrastructure to integrate
quantum resources into HPC applications. OpenMQPI introduces QPIR, a lightweight MLIR dialect for representing
quantum circuits within OpenMP target regions, and a compiler pipeline that performs
quantum-specific optimizations and lowers
circuits to executables compatible with OpenMP offloading. At runtime, an MQSS plugin enables the generated executable to offload
quantum circuits to
quantum hardware through the OpenMP target offloading model. We evaluate OpenMQPI using the VQE, QFT, and Deutsch-Jozsa
algorithms with 5, 10, 15, and 20
qubits, comparing execution times against Qiskit and CUDA
Quantum. OpenMQPI achieves the lowest average VQE execution time across all tested
qubit configurations. At 20
qubits, it requires 5.18 s, compared with 5.83 s for Qiskit and 7.39 s for CUDA
Quantum. For QFT and Deutsch-Jozsa, OpenMQPI achieves execution times comparable to Qiskit and lower than those of CUDA
Quantum.
Read the full abstract ▾
All authors Ercüment Kaya, Ekin Devim Şahinkaya, Minh Chung, Martin Schulz
Machine Learning · Cross-listed from stat.ML
Reliable Scale-Free Bandits via Empirical Likelihood
Pratham Gupta, Alhad Sethi and Shubhada Agrawal
We study
stochastic bandits without knowledge of reward scales or tail parameters. Almost-sure regret guarantees describe eventual learning but do not quantify expected performance or the risk of large regret in a finite run. We introduce $\texttt{EL-F}$, an anytime empirical-likelihood (EL) extension of the $g$-Forcing policy of Cowan and Katehakis (2020), and characterize its performance at three levels. First, for integrable rewards and non-decreasing, diverging exploration schedules $g$ with vanishing increments, $\texttt{EL-F}$ preserves the $g$-Forcing asymptotics for pseudo-regret $R(T)$: $R(T)/g(T)\to\sum_{i:\Delta_i>0}\Delta_i$ almost surely as $T\to\infty$, where $\Delta_i$ is the suboptimality gap of arm $i$. Second, on every sub-exponential instance, for non-decreasing $g(t)=\omega(\log t)$ satisfying a mild feasibility condition, expected pseudo-regret is at most $g(T)\sum_{i:\Delta_i>0}\Delta_i+C_{\underline{P}}$, where $C_{\underline{P}}$ is an instance-dependent constant independent of $T$. Moreover, we also establish finite-time tail bounds with super-
polynomial decay in regret beyond the forced-exploration budget. For slowly varying $g$, we determine the exact exponent of linear regret deviations on two-armed Gaussian instances with common variance and establish a matching lower bound among policies satisfying the same expected-regret budget on every such Gaussian instance. These quantitative guarantees also cover the greedy variant obtained at zero EL radius. A fixed positive radius introduces local optimism while preserving the leading asymptotic regret. We establish coupled pathwise dominance over the greedy variant in two-armed bandits with a deterministic suboptimal arm. Experiments demonstrate reductions in prolonged suboptimal allocation compared with $g$-Forcing on challenging instances.
Read the full abstract ▾
Cryptography and Security
Privacy-Preserving GNN Inference on Financial Transaction Graphs via Homomorphic Encryption
H. Çağr{\i} Bilgi, Kemal Mutluergil, Ceren Y{\i}ld{\i}r{\i}m et al. (+2 more)
Anti-money laundering (AML) in financial networks requires modeling transaction patterns that span multiple institutions. In practice, however, banks cannot share account and transaction features due to strict
privacy regulations. We propose a
privacy-preserving graph neural network (GNN) inference framework tailored to this multi-bank setting. A central server holds the graph connectivity and pretrained model weights in plaintext, while each bank keeps its transaction features encrypted. Inference is then carried out entirely on encrypted features under the CKKS homomorphic encryption scheme, so that no plaintext transaction features or predictions are exposed to the server. The key to making this practical is an edge-centric GNN architecture, wherein message passing is expressed through operations on node–edge incidence
matrices. This formulation not only preserves the edge-attributed multigraph structure of transaction networks but also reduces every aggregation step to plaintext–ciphertext
matrix products. Since CKKS can evaluate only
polynomial circuits, we also approximate each nonlinearity with a composite
polynomial of the sign function, which we calibrate per activation site. On the IBM AML
benchmark, our model achieves performance competitive with plaintext baselines, and with a consumer-grade GPU, classifies each queried transaction end-to-end in as little as 5.31 seconds, while per-transaction latency can be amortized to 1.65 seconds when optimized for throughput.
Read the full abstract ▾
All authors H. Çağr{\i} Bilgi, Kemal Mutluergil, Ceren Y{\i}ld{\i}r{\i}m, Kubilay Atasu, Erkay Savas
Robotics
LADDER: A Long-Horizon Robot Manipulation Benchmark
Petko Petkov, Giuliano Albanese, Nikolay Nikolov et al. (+3 more)
General-purpose
robot manipulation policies learn short-horizon skills effectively, but degrade rapidly on long-horizon tasks. Existing
benchmarks cannot isolate failures because task horizon is either too short or entangled with scene and object variations. We propose LADDER, a long-horizon
robot manipulation
benchmark scaling from 1 to 10 subgoals within the same environment over three suites (SimplerEnv-Long, RobotArena-$\infty$-Long, and RoboLab-Long), 145 tasks, eight categories, two embodiments, and 2,024 simulated episodes per system. We evaluate two orchestration paradigms: text subgoals for learned low-level policies (VoLo) and tool selection orchestration (CaP-X, GPT-Policy, PSP-
Agent, TiPToP). Given the full instruction directly, low-level policies achieve 85.4% on single-subgoal tasks, but drop to near zero on extended tasks. While orchestration remedies this, it remains bounded by physical execution errors of the low-level policy rather than high-level planning and achieves only ~25%. Tool selection orchestration with GPT-Policy on GPT-6 Astra significantly outperforms learned-policy methods, but incurs significant time and cost budget and succeeds only ~62%. Real-world experiments across two embodiments confirm the collapse of the low-level policy and the gain from orchestration. Project website: https://ladder.insait.ai
Read the full abstract ▾
All authors Petko Petkov, Giuliano Albanese, Nikolay Nikolov, Luc Van Gool, Jonathan Tremblay, Danda Pani Paudel
Artificial Intelligence
Bofan Xie, Chaokun Wang, Guanchen Ge et al. (+4 more)
Long-term
memory has become a core component of large
language model (LLM)
agents. As
agents accumulate personal information over long-term use, users increasingly ask them to delete sensitive or outdated facts. A
memory system should honor such requests. It should remove exactly what was requested and keep the rest of its
memory intact. However, existing
memory benchmarks evaluate only
memory writing,
retrieval, and organization, so whether
agents can forget at the user's explicit request remains under-evaluated. To address this gap, we propose Bemula, the first
benchmark dedicated to
memory unlearning. It comprises 50 multi-session dialogue instances, 135 forget requests of six types, and 1,742 test questions. Each dialogue has a control version without the requests, so that only systems that have stored the target can earn credit for forgetting. Across the paired versions, five test suites measure storage, removal of the requested information, preservation of unrelated
memories, relearning of similar facts, and resistance to extraction attacks. On Bemula, we evaluate six mainstream
memory systems under three backbone LLMs, together with the Full Context LLM on two of them. The results are negative. Across all evaluated systems and backbones, no system passes more than 1.5% of forget requests end to end, and most pass none at all. These systems either leave the requested information extractable under simple attacks or delete far more than requested.
Memory unlearning therefore remains an open challenge for
agent memory design. We release our data and code at https://github.com/Nboxff/bemula-
benchmark.
Read the full abstract ▾
All authors Bofan Xie, Chaokun Wang, Guanchen Ge, Songyao Wang, Fang Niu, Hang Zhang, Wei Wang
Machine Learning
inLoRA-MINT: Internal Representations for Auditing Training Data in Domain-adapted LLMs
Gonzalo Mancera, Julian Fierrez, Aythami Morales et al. (+3 more)
This work investigates the application of the Membership Inference Test (MINT) to large
language models (LLMs) fine-tuned with LoRA, aiming to determine whether specific samples were used during training and address concerns about data
privacy, leakage, and transparency. Building on our previous LoRA-MINT framework, which relies exclusively on perplexity under a strict black-box setting, we propose an approach based on intermediate model representations, including activations and parameter-dependent signals. These representations enable a MINT classifier to distinguish training from non-training samples without requiring full access to model parameters. We evaluated the method on four LLMs (Qwen, Phi-4, DeepSeek-V3, and LLaMA 3.2-3B) fine-tuned on three
datasets covering medical, mathematical, and academic
reasoning tasks. Results show consistent performance across architectures and domains, achieving scores between 95% and 98%, with robustness to variations in model configurations and data distributions. The proposed approach enables efficient auditing of sensitive training data and supports data provenance and compliance with emerging
privacy and AI governance regulations. Although our experiments focus on LoRA-based fine-tuning, the methodology can be extended to other LLM adaptation techniques and, more broadly, to other AI models.
Read the full abstract ▾
All authors Gonzalo Mancera, Julian Fierrez, Aythami Morales, Ruben Tolosana, Daniel DeAlcala, Francisco Jurado
Artificial Intelligence
Beyond Black Boxes: Visual Grounding for Text Embeddings
Slimane Larabi
In this paper, our goal is to represent text
embeddings as sketches for three main reasons. First, to visualize text representation and to explore their potential their potential for AI explainability. Second, to investigate the potential benefits of the sketch-based representation for information
retrieval. Third, to explore their potential for
reasoning, drawing inspiration from the concept of mental imagery. We propose a sketch-based visual grounding framework that transforms text representations into visual topographic sketches. Our approach aims to emphasis the semantics content encoded in the text
embeddings through the corresponding sketch. Evaluations are conducted on the MS-COCO
dataset to examine the utility of these sketches for text
retrieval. The comparative study of SBERT and Sketch-based method reveals that SBERT retains an advantage in strict top-1
retrieval accuracy while achieving comparable performance to the sketch-based method in top-5
retrieval. A cascaded approach combining SBERT and the sketch-based method outperforms SBERT alone, suggesting that the Sketch-based representation could offer significant potential for information
retrieval if trained directly on text.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from physics.flu-dyn
High-fidelity Three-Dimensional Flow Field Reconstruction Driven by Turbulent Fluid Dynamics
Hanlin Kong, Jianan Wei, Yuntian Chen et al. (+2 more)
High-fidelity three-dimensional turbulent flow fields are essential for scientific and engineering analyses, yet obtaining them through direct numerical simulation or high-resolution measurements remains expensive. Super-resolution reconstruction from coarse-grained data therefore offers a promising alternative, but remains fundamentally challenging because fine-scale structures are intrinsically three-dimensional and dynamically coupled across scales. Existing data-driven approaches still struggle to recover multiscale volumetric structures and their coherent temporal
evolution. To address these limitations, we propose a
turbulence-aware framework for time-resolved three-dimensional flow reconstruction. The framework couples a multiscale neural operator, designed to recover volumetric structures, with a state space model to promote temporal consistency in the reconstructed dynamics. In addition, Navier-Stokes residuals are incorporated as soft constraints to regularize the reconstruction toward physically plausible dynamics. A mixture-of-experts mechanism is also introduced to support unsupervised adaptation to unknown coarse-graining. Across evaluations on canonical turbulent systems and atmospheric reanalysis data, the proposed framework reduces reconstruction errors relative to learning-based and classical baselines, while better preserving multiscale spatial structure and coherent temporal
evolution. These results validate the effectiveness of the proposed framework in reconstructing high-fidelity three-dimensional flow fields from coarse-grained data.
Read the full abstract ▾
All authors Hanlin Kong, Jianan Wei, Yuntian Chen, Yujia Wang, Wenguan Wang
Robotics
Contractor Filters
Simon Rohou
We introduce Contractor Filters (CFs), a set-membership architecture for nonlinear continuous-time state estimation with bounded uncertainties and sporadic outliers. CFs represent feasible states symbolically by compositions of contractors rather than rematerializing a
geometric enclosure after each update. This provides an abstraction layer between the set-theoretic filtering equations and their numerical realization, so prediction, correction, ambiguity, and outlier handling can be expressed within one
algebra without prescribing a state geometry. We derive an exact nominal recursion and a guaranteed robust enclosure, with validated flow inclusions preserving soundness. Optional sub-paving compression then provides finite outer representations. CFs are evaluated against a robust particle filter on range-only and terrain-based localization under outliers, probabilistic model mismatch, association ambiguity, and multimodality. CFs retain the simulated truth in all tested configurations, while particle-filter reliability degrades under persistent bias and spatial aliasing. Contractor strength and compression budget control the trade-off between paving tightness and computational cost, while preserving the enclosure guarantee.
Read the full abstract ▾
Computer Vision
Scaling Vision Transformers Without Training: Prune-and-Merge Token Compression
Vatsal, Soumil Malhotra, Priyanshu Chahal et al. (+1 more)
Vision Transformers achieve strong performance but incur high computational costs from quadratic
token interactions. Training-free
token reduction methods improve efficiency, but their robustness remains underexplored. We propose Prune-and Merge ViT, a training-free framework for off-the-shelf models. It combines class
attention-based
pruning with similarity-based merging to reduce redundancy while preserving semantic information. To evaluate robustness we introduce two metrics. Relative Accuracy Drop (RAD) measures relative accuracy degradation under transformations, and Prediction Flip Rate (PFR) quantifies prediction consistency. On pretrained ViT and DeiT, Prune-and-Merge reduces computation competitively while maintaining robustness and better accuracy than existing
token reduction methods. Our analysis suggests that hybrid
token reduction does not introduce
hallucination-like behavior and preserves prediction stability for common image transformations.
Read the full abstract ▾
All authors Vatsal, Soumil Malhotra, Priyanshu Chahal, Sanjeev Kumar
Robotics
Data-Free Continual Learning for World Action Model with Rectified Generative Replay
Kaipeng Fang, Hangxu Liu, Chaojian Shi et al. (+5 more)
World action models (WAMs) built on pretrained
video generation models adapt quickly to new robotic tasks, but they forget previously learned tasks under sequential fine-tuning. Experience replay (ER) mitigates such forgetting by rehearsing data of previous tasks, yet this data is often unavailable in robotics, where a policy is deployed as weights only and its training data is not shared along with it. In this work, we ask whether a WAM can achieve data-free continual learning by regenerating its own past. Our key observation is that a WAM can indeed autoregressively regenerate complete rollouts of previously learned tasks and replay them to mitigate forgetting. However, replaying such raw rollouts still lags far behind replaying real data, which we attribute to two issues. First, the WAM is unable to recognize task completion and keeps generating purposeless motion after the task is done. Second, under open-loop generation, the generated actions gradually drift away from the generated observations. To address both issues, we present rectified generative replay (R$^2$-WAM), which truncates generated rollouts at task completion with an agentic evaluator and filters out rollouts whose actions are physically inconsistent with the observations. On the four LIBERO suites and five real-world tasks, R$^2$-WAM reduces forgetting by over 90% relative to sequential fine-tuning and nearly matches ER without accessing any data from previous tasks. Our analysis further reveals that forgetting in WAMs concentrates in inverse dynamics modeling: a forgotten WAM still produces visually plausible observations of previous tasks while its generated actions fail. Project page: https://kaipengfang.github.io/r2-wam/
Read the full abstract ▾
All authors Kaipeng Fang, Hangxu Liu, Chaojian Shi, Yunsong Zhou, Hongzi Zhu, Minyi Guo, Jingkuan Song, Lianli Gao
Machine Learning · Cross-listed from stat.ML
Weighting-Based Strategies for Classification: Analysis and Local Peak Dynamics with CART Trees
Abdoulaye Sakho, Emmanuel Malherbe and Erwan Scornet
Tabular learning tasks often suffer from class imbalance, which is often addressed via weighting-based strategies. In this work, we compare the naive approach (unweighted) with the weighting-based rebalancing strategies. Somewhat surprisingly, we demonstrate that, asymptotically, as the predictor complexity increases with sample size, weighting strategies yield predictive performance identical to unweighted approaches across all evaluation metrics, up to a shift in the final threshold. To go beyond, we theoretically study the estimation of a linear regression function through weighted and unweighted CART trees with limited depths. In this setting, we prove that the points in the PR space for both strategies are distributed along the same curve, thus illustrating the similar nonasymptotic performance of the two strategies. Then, we numerically identify more complex synthetic settings where performance gaps, evaluated via fixed classifier complexity, favor weighting strategies. Thus, to guide practitioners, we introduce an indicator that aims at signaling whether a data set may benefit from weighting.
Read the full abstract ▾
Computer Vision
COMPASS: Contextual Moment Prior Alignment for Generalizable AIGI Detection
Zhan Yang, Tiancheng Shen, Weijie Lyu et al. (+4 more)
Generalizable AI-generated image (AIGI) and deepfake detection remain challenging due to the asymmetry phenomenon, where adaptation to limited training distributions can degrade the transferable representations of
Vision Foundation Models (VFMs). Recent subspace adaptation methods mitigate this issue by freezing high-magnitude principal components while adapting the residual subspace. However, such magnitude-based decomposition is data-agnostic and does not explicitly account for target-relevant semantics, potentially leading to semantic misalignment. We propose COMPASS (COntextual Moment Prior
Alignment for Subspace Separation), a data-aware subspace adaptation framework that uses a small unlabeled Semantic Proxy Set to estimate target activation statistics. COMPASS incorporates activation covariance into the decomposition of pre-trained weights, preserving target-relevant representations as a stable semantic reference while the residual adapter learns forgery-related variations. The proxy calibration requires no gradient-based optimization. Extensive experiments demonstrate strong generalization across cross-
dataset, cross-method, joint cross-
dataset-method, and open-world cross-generator evaluations, achieving state-of-the-art performance on both Deepfake and AIGI detection
benchmarks.
Read the full abstract ▾
All authors Zhan Yang, Tiancheng Shen, Weijie Lyu, Shuo Chen, Bernard Ghanem, Yibo Yang, Ming-Hsuan Yang
Computer Vision
FlowSim: Rendering-Free Closed-Loop Simulation for Evaluating End-to-End Driving Systems
Hanfeng Wu, Alvaro Marcos-Ramiro, Michael Schmidt et al. (+1 more)
Closed-loop evaluation of end-to-end (E2E) driving systems typically relies on generating sensor observations and re-processing them through the perception encoder, introducing substantial computational cost and a sim2real gap between simulated and real sensor observations. In this paper, we ask a more fundamental question: is image-space simulation necessary for closed-loop evaluation of E2E driving systems? We propose FlowSim, a rendering-free simulation paradigm that predicts the internal latent representation directly from a previous latent and target-frame symbolic scene conditions. The predicted latent is then processed by the original frozen downstream modules, eliminating image rendering and perception re-encoding. Explicit cross-frame
agent association improves latent transition prediction and rollout stability, while variable-horizon simulation enables long-horizon prediction from a common real anchor without accumulating intermediate rollout errors. Experiments on nuScenes and NavSim demonstrate that FlowSim preserves downstream behavior competitively with, and often exceeds image-space simulation in source-to-reference agreement across variable simulation horizons and both dense and sparse latent representations. By predicting the latent directly, both instantiations of FlowSim simulate each step in 0.04–0.06s, substantially faster than the image-space simulation methods they are compared against. These results establish FlowSim as an efficient and practical alternative to conventional image-space simulation methods.
Read the full abstract ▾
All authors Hanfeng Wu, Alvaro Marcos-Ramiro, Michael Schmidt, Christoph Stiller
Artificial Intelligence · Cross-listed from physics.chem-ph
Can Machine Learning Potentials Be More Accurate Than the Methods They Are Trained On?
Pavlo O. Dral, Yi-Fan Hou, Yuxinxin Chen et al. (+4 more)
Although machine learning potentials (MLPs) generalize across configurational and chemical compound spaces, their accuracy is commonly believed to be limited by their training methods, with improvements beyond this limit achieved so far only by refitting to experimental observables. In this Perspective, we argue that MLPs can also be made more accurate than their training methods non-empirically by generalizing across a third space, that of
quantum-chemical methods. We propose routes to method-transferable MLPs that predict at unseen methods, with working implementations in both density-functional and wavefunction-based theories showing that they can be more accurate than any of their training methods. Aiming beyond the accuracy of the training methods could also make universal MLPs more accurate than single-method models, even when the method-transferable MLPs remain less accurate than those QM methods. Method-transferable MLPs will shape the construction of training
datasets and the evaluation of models, pushing the accuracy limits of chemical simulations.
Read the full abstract ▾
All authors Pavlo O. Dral, Yi-Fan Hou, Yuxinxin Chen, Hassan Nawaz, Mateusz Witkowski, Szymon Śmiga, Ireneusz Grabowski
Machine Learning
ChampionKV: Head-Wise Adaptive Budget Allocation for KV Cache Compression
Yuechuan Chen, Yanda Guo, Xiaoxi Liu et al. (+4 more)
Some observation-window-based KV cache compressors use
prompt attention to identify important
tokens but allocate uniform or predefined budgets across heads, overlooking input-dependent differences in budget demand. We propose ChampionKV, an input-adaptive head-wise budget allocator for long-context KV cache compression. It uses the maximum aggregated observation-window
attention score as a lightweight empirical proxy for
attention concentration, assigning larger budgets to heads with dispersed
attention and smaller budgets to concentrated heads. An abnormal-head mask and a floor-budget mechanism improve robustness under weak
attention responses. ChampionKV plugs into compressors such as SnapKV and PyramidKV without changing their
token-selection procedures. On LongBench and RULER, both Champion variants consistently improve over their SnapKV and PyramidKV base compressors; the improvements are largest and statistically significant on Llama3.1-70B-Instruct, especially under tight budgets, whereas results on the tested 7B/8B models remain competitive but mixed. ChampionKV does not uniformly outperform Ada-KV, whose global cross-head Top-K selection retains an advantage in several settings. Controlled analysis further shows improved retained
attention mass over the corresponding base compressors at negligible overhead.
Read the full abstract ▾
All authors Yuechuan Chen, Yanda Guo, Xiaoxi Liu, Weilin Ruan, Huanbo Xu, Ruyi Wang, Wei Zhu
Artificial Intelligence
SHELDON: Reasoning through Dual Search over Hypothesis and Observation Spaces
Harshit Joshi, Priyank Shethia and Monica S. Lam
Test-time scaling has enabled LLMs to solve increasingly complex
reasoning, discovery, and optimization tasks. However, existing approaches often gather observations while seeking evidence for current hypotheses, then generate new hypotheses conditioned on those hypotheses and their associated evidence. This couples observation discovery to existing hypotheses, favoring local refinements over exploration of unknown hypotheses regions. Inspired by the hypothetico-deductive method (Hempel, 1945) and dual-space models of scientific discovery (Klahr & Dunbar, 1988), we introduce SHELDON, a scientific
reasoning harness that separately represents and jointly explores hypothesis and observation spaces. By actively seeking new observations along with new hypotheses, SHELDON can discover observations that its current hypotheses do not explain. These unexplained observations expose gaps in its understanding, motivate new investigations, and guide the generation of hypotheses beyond refinements of existing ones. Across diagnosis, abstract
reasoning, and optimization
benchmarks, SHELDON outperforms evaluated baselines by up to 26.7 points. Additional analysis shows that SHELDON generates up to 4.5x more novel solutions.
Read the full abstract ▾
Machine Learning
FRACTION: Fraud Explanation Generation via Agentic Graph Context Navigation
Cecelia Soh, Jun Hu, Yuan Li et al. (+6 more)
Graph fraud explanation generation emerges as an important task, which requires jointly leveraging graph structure and content information to produce human-readable explanations for a given risk label. An intuitive approach is to adapt existing graph-enhanced large
language models (LLMs), which inject graph structure into LLMs via graph-conditioned instruction tuning or structure-aware
embeddings. However, existing methods face two key challenges in graph fraud explanation generation: a lack of evaluation protocols tailored to explanation quality, and a mismatch between the granularity of graph information they provide and what explanation generation requires. In this paper, we propose FRaud Explanation Generation via Agentic Graph Context NavigaTION (FRACTION), a framework for graph fraud explanation generation. To evaluate explanation quality, we introduce a task-specific rubric that assesses faithfulness, graph grounding, helpfulness, consistency, and conciseness, and validate it against human judgments. To focus on fine-grained details of relevant neighbors without reading the entire neighborhood, we equip a
language model with graph tools that progressively disclose neighbor records. To train our graph
agent, we propose a rubric-guided trajectory
distillation strategy through which the
agent learns both tool-use behavior and structured explanation generation. Experimental results on three fraud
datasets demonstrate consistent improvements over state-of-the-art baselines under a human-aligned rubric.
Read the full abstract ▾
All authors Cecelia Soh, Jun Hu, Yuan Li, Cheng Chen, Hui Liu, Yanjia Zhu, Hao He, Bryan Hooi, Bingsheng He
Information Theory
The Gauss–Newton Method for Phase Retrieval from Coded Diffraction Patterns
Meng Huang and Tianjiao Wang
We study the minimum-norm Gauss–Newton method for phase
retrieval from random coded diffraction patterns with additive intensity noise. For bounded symmetric masks satisfying the standard complex moment conditions, we establish convergence guarantees with quadratic error recursions in both the resampled and fixed-data settings. In the resampled setting, $O(\log^2 n)$ masks per independent block suffice to obtain a spectral initializer within a constant-relative-radius neighborhood and to establish a local quadratic error recursion with an additive noise term. Consequently, in the noiseless case, the method achieves relative accuracy $\epsilon$ using $O(\log\log(1/\epsilon)\log^2 n)$ masks. In the fixed-data setting, $O(\log^6(en))$ masks suffice for spectral initialization within a neighborhood whose relative radius is proportional to $\log^{-5/2}(nL)$. Within this neighborhood, the method satisfies a linear contraction bound and eventually a quadratic error recursion. In particular, in the noiseless case, the method initially converges linearly and enters a quadratic convergence phase after $O(\log n)$ iterations, without acquiring additional masks. Numerical experiments on synthetic signals and a natural image demonstrate the method's efficiency relative to state-of-the-art
algorithms.
Read the full abstract ▾
Computation and Language
Machine–human agreement in marking undergraduate assignments is specific to model, prompt and institution
Giulio Corsi, Maryam Abo-Tabik, Yael Benn et al. (+4 more)
Large
language models (LLMs) promise to automate one of higher education's most resource-intensive activities: assessing student writing. Existing evaluations are typically limited to a single model, a single
prompt and a single institution. Using 761 undergraduate psychology assignments from three UK universities, we evaluated three frontier LLMs, individually and in ensemble, on accuracy,
prompt sensitivity, error independence, generalisability and evidential basis. Agreement between human and machine marks varied widely, with no model or
prompt optimal across institutions. All models compressed marks towards a common central tendency, overvaluing the weakest assignments and undervaluing the strongest. Because this error was shared, ensembling failed to remove it. Machine marks also rewarded linguistic features more strongly than human marks. Since human-machine agreement was a joint product of model,
prompt and institution, we conclude that current LLMs lack evaluative capacity transferable across contexts, and that single-context validation overstates readiness for deployment.
Read the full abstract ▾
All authors Giulio Corsi, Maryam Abo-Tabik, Yael Benn, Georgiana Thorpe Apreutesei, Roni Tibon, Alexandru Marcoci, Deborah Talmi
Computer Vision
Dynamic Interpolation for Training-Time Prediction Reshaping in Low-Light Image Enhancement
Peirui Zhu, Shijie Hao, Jingxi Liao et al. (+2 more)
Low-light image enhancement (LLIE) is an important image/
video processing problem for reliable visual systems under poor illumination, yet it remains challenging due to the strong coupling among insufficient illumination, noise amplification, and structural degradation. Existing methods typically rely on sophisticated network architectures or engineered objective functions. While effective, the former increases model complexity and deployment burden, whereas the latter relies on empirically tuned loss weights. We revisit supervised LLIE from a training-dynamics perspective and observe that brightness-related errors tend to converge early, while residual noise and structural artifacts often remain insufficiently optimized in later stages. To address this issue, we propose a Dynamic Interpolation Training-only Module (DITM), which introduces no additional learnable parameters and adaptively reshapes the model prediction for supervised feed-forward LLIE training. Specifically, DITM employs brightness-aware interpolation, in which the model prediction is combined with a mean brightness-aligned image constructed by rescaling the model prediction according to the mean brightness of the ground-truth image. By reducing the relative dominance of global brightness discrepancy, DITM makes residual local errors more influential during backpropagation after global luminance has largely stabilized, thereby facilitating artifact suppression and fine-detail restoration in later training stages. DITM is used only during training and is fully removed at inference, introducing no additional inference cost. Extensive experiments on multiple
benchmark datasets and CNN-,
Transformer-, and Mamba-based LLIE backbones show that DITM consistently improves fidelity and perceptual quality while preserving the inference-time architecture and introducing no additional inference-time computation.
Read the full abstract ▾
All authors Peirui Zhu, Shijie Hao, Jingxi Liao, Richang Hong, Meng Wang
Data Structures and Algorithms
Inclusion-wise extremal knapsack problems are (mostly) hard to approximate
Fabio Ciccarelli, Fabio Furini, Ulrich Pferschy et al. (+1 more)
We study two inclusion-wise extremal variants of the 0-1 Knapsack Problem (KP). Given $n$ items with integer weights and nonnegative costs or profits, the Minimum-Cost Maximal Knapsack Packing Problem seeks a minimum-cost packing for capacity $c$ to which no item can be added, while the Maximum-Profit Minimal Knapsack Cover Problem seeks a maximum-profit cover for threshold $d$ from which no item can be removed. Both problems are weakly $\mathrm{NP}$-hard and pseudo-polynomially solvable, as is the classical KP, but their approximability is fundamentally different: unless $\mathrm{P}=\mathrm{NP}$, with positive values neither problem admits a
polynomial-time $2^{p(n)}$-approximation for any
polynomial $p$; allowing zero values rules out every multiplicative factor. Even with unit costs for packing and unit profits for covering, neither admits an $n^{1-\varepsilon}$-approximation for any fixed $\varepsilon>0$, whereas a trivial $n$-approximation exists. We identify restrictions that restore approximability. If every item weighs at least a fixed fraction of the capacity, respectively of the threshold, both problems can be solved exactly in
polynomial time. For the cover problem, if each weight $w_i$ satisfies $w_i-1\ge\delta(d-1)$, we give an approximation scheme for arbitrary profits whose running time is
polynomial in $n$, $1/\varepsilon$, and $1/\delta$. If values equal weights, a fully
polynomial-time approximation scheme (FPTAS) was already known for the packing problem; we give an FPTAS for the cover problem, improving on the previously known factor $2$. If value-to-weight ratios differ by at most a factor $\kappa$, we obtain a $(1+\varepsilon)\kappa$-approximation for both problems.
Read the full abstract ▾
All authors Fabio Ciccarelli, Fabio Furini, Ulrich Pferschy, Elena Rener
Computation and Language
LatentRemedy: Efficient Remedy For Non-Prefix KV Cache Reuse in Multi-Agent Systems
Zhiwei Chen, Zhongtao Rao, Waikit Xiu et al. (+4 more)
Non-prefix KV cache reuse is a promising approach for improving inference efficiency in multi-
agent systems as
agents share the same context blocks. However, when
agents have conflicting tasks, directly reusing non-prefix KV caches across
agents can severely degrade model performance, causing up to a 90% drop. Existing remedies, including selective recomputation and offset-based calibration, recover a limited fraction of this loss. We argue that the problem lies not only in contaminated KV representations but also in the
agent's difficulty interpreting reused context under its own instructions. To this end, we introduce LatentRemedy, which uses a few task-conditioned latent
tokens to help the target
agent leverage reused context without recomputing the shared block. Specifically, we use an MLP to map the last hidden state of the target prefix into a small number of latent
tokens, which are placed after the reused KV block to interact with the shared context. To guide subsequent predictions toward the target task, we train the MLP through full-prefill
distillation. Experiments show that LatentRemedy consistently outperforms existing reuse baselines on conflicting tasks while maintaining TTFT close to direct reuse as cache size grows. Code and data are available at https://github.com/chenzhwsysu57/kvreuse.
Read the full abstract ▾
All authors Zhiwei Chen, Zhongtao Rao, Waikit Xiu, Yuxiang Zhang, Shuting Cui, Jian Chen, Tianxiang Zhao
Machine Learning
Correcting or Continuing? What On-Policy Distillation Learns from Student Errors
Shuo Yang, Zheyu Zhang, Mahito Sugiyama et al. (+1 more)
On-policy
distillation (OPD) trains a student on its own rollouts under a frozen teacher, mitigating the distribution mismatch of off-policy supervision. While prior work has primarily trained on problems a student can largely solve, we ask whether OPD can improve a student when training is restricted to problems it currently fails. We evaluate five representative OPD mechanisms with Qwen-3 and Phi-4 models under a unified competition-mathematics protocol. We find that OPD does not improve accuracy on this hard subset, with most mechanisms instead degrading performance despite continued decreases in training KL divergence. All eight configurations show a decline, and the two-sided sign test gives p=0.0078. In contrast, the same pipeline improves accuracy by 4.6 percentage points when training is not concentrated on failures, showing that the failure is conditional on the training regime rather than systematic. These findings suggest that OPD is effective when supervision is dominated by correct student prefixes, but can reinforce existing errors when training is concentrated on problems the student currently fails.
Read the full abstract ▾
All authors Shuo Yang, Zheyu Zhang, Mahito Sugiyama, Gjergji Kasneci
Sound
Fish Audio Drama 3 Technical Report
Dawei Han, Shidong Li, Shijia Liao et al. (+13 more)
We introduce Fish Audio Drama 3, an open-source
speech generation model that turns text into directed performances, built on a 35B-parameter mixture-of-experts backbone that activates 3B parameters per
token. It clones voices from multiple references, follows natural-
language direction, revises a take when the next turn gives a note, voices multi-character scenes, edits existing
speech by infilling, and synthesizes long-form and streaming audio. Multi-turn revision, scene dubbing, editing by infilling, and streaming text input are new in Drama 3, all in one model. Drama 3 pairs the hybrid linear-
attention backbone with a new codebook multi-
token-prediction audio decoder that predicts the acoustic codebooks of each frame in dedicated layers that attend over the entire sequence. The model is trained with a three-stage pre-training recipe, instruction fine-tuning, and large-batch reinforcement learning whose only reward is a single learned model of human preference. Its inference engine decodes a single stream at 271 frames per second on one H200, against 177 for Fish Audio S2, although Drama 3 has more than seven times as many parameters. The model weights are available on GitHub (https://github.com/fishaudio/fish-
speech) and Hugging Face (https://huggingface.co/fishaudio/drama-3), and inference is available through SGLang-Omni (https://github.com/sgl-project/sglang-omni). We highly encourage readers to visit https://fish.audio to direct their own performances.
Read the full abstract ▾
All authors Dawei Han, Shidong Li, Shijia Liao, Jiahua Liu, Songting Liu, Qingzheng Wang, Yuxuan Wang, Xin Chen, Yifan Cheng, Tianyu Li, Gangin Park, Xuehao Yang, Chang Zeng, Haoyang Zhang, Ruoyi Zhang, Xiangyu Zhang