Machine Learning
Modular Multimodal Classification Without Fine-Tuning: A Simple Compositional Approach
Herman Bergström, Aditya Mehrotra and Rahul G. Krishnan
We introduce CoMET, \textit{Composing Modality Encoders with Tabular foundation models}, a simple yet highly competitive method for
multimodal classification: pass each modality through a frozen pre-trained backbone, compress the resulting
embeddings with PCA, and concatenate as input into a Tabular Foundation Model (TFM) for prediction. We show that PCA alone suffices to act as an adaptor yielding strong, robust performance across modalities. When the CLS
tokens of the foundation model align poorly with downstream tasks, we propose PALPooling, a lightweight adaptive
token pooler that consistently improves representation quality. By composing strong frozen representation learning backbones with TFMs, our approach achieves state-of-the-art results across diverse
multimodal benchmarks without any training. On hierarchical tasks with large fine-grained class spaces, our approach enables fast and scalable classification, handling
datasets with over 500,000 samples and 2,000 classes without any fine-tuning. Overall, our results show that the composition of foundation models is a simple, yet powerful, out-of-the-box solution for
multimodal learning, challenging the necessity of complex, end-to-end training pipelines for new problems.
Read the full abstract ▾
Machine Learning
Chunyuan Deng, Yizhe Zhang, Rui-Jie Zhu et al. (+4 more)
Looped
Transformers (LT) have emerged as a powerful architecture by iterating their layers multiple times before decoding the final
token. However, pairing them with full
attention retains quadratic complexity, making them computationally expensive and slow. We introduce LT2 (Linear-Time Looped
Transformers), a family of looped architectures that replace quadratic softmax
attention with subquadratic, linear-time
attention. We study two variants: LT2-linear with linear
attention and LT2-sparse with sparse
attention. We find that looping uniquely synergizes with these variants: it enables iterative
memory refinement in linear
attention and progressively expands the effective receptive field in sparse
attention. We formalize these benefits theoretically and demonstrate consistent empirical gains across controlled recall, state-tracking, and
language modeling tasks. We then explore LT2-hybrid, which combines different
attention variants in a looped setting. Two variants are especially promising: LT2-hybrid (GDN+DSA), which interleaves linear and sparse
attention to maximize efficiency and matches the standard looped
transformer's quality at fully linear-time cost; and LT2-hybrid (Full+GDN), which interleaves GDN with a small fraction of full
attention layers to maximize quality, surpassing the standard looped
transformer in both performance and efficiency. We also show how to convert a pre-trained LT into an LT2-hybrid model. With about 1B
tokens of training, our converted model, Ouro-hybrid-1.4B, outperforms industry-level 1B models and is competitive with industry-level 4B models while retaining the speed benefits of linear-time
attention. Together, these results show a clear path toward making looped
transformers more scalable and advancing efficient, capable small
language models.
Read the full abstract ▾
All authors Chunyuan Deng, Yizhe Zhang, Rui-Jie Zhu, Yuanyuan Xu, Jiarui Liu, T. S. Eugene Ng, Hanjie Chen
Machine Learning · Cross-listed from cs.CL
On the limits and opportunities of AI reviewers: Reviewing the reviews of Nature-family papers with 45 expert scientists
Seungone Kim, Dongkeun Yoon, Kiril Gashteovski et al. (+55 more)
With the advancement of AI capabilities, AI reviewers are beginning to be deployed in scientific peer review, yet their capability and credibility remain in question: many scientists simply view them as probabilistic systems without the expertise to evaluate research, while other researchers are more optimistic about their readiness without concrete evidence. Understanding what AI reviewers do well, where they fall short, and what challenges remain is essential. However, existing evaluations of AI reviewers have focused on whether their verdicts match human verdicts (e.g., score
alignment, acceptance prediction), which is insufficient to characterize their capabilities and limits. In this paper, we close this gap through a large-scale expert annotation study, in which 45 domain scientists in Physical, Biological, and Health Sciences spent 469 hours rating 2,960 individual criticisms (each targeting one specific aspect of a paper) from human-written and AI-generated reviews of 82 Nature-family papers on correctness, significance, and sufficiency of evidence. On a composite of all three dimensions, a reviewing
agent powered by GPT-5.2 scores above each paper's top-rated human reviewer (60.0% vs. 48.2%, p = 0.009), while all three AI reviewers (including Gemini 3.0 Pro and Claude Opus 4.5) exceed the lowest-rated human across every dimension. AI reviewers' accurate criticisms are also more often rated significant and well-evidenced, and surface a distinct 26% of issues no human raises. However, AI reviewers overlap far more than humans do (21% vs. 3% for cross-reviewer pairs), and exhibit 16 recurring weaknesses humans do not share, such as limited subfield knowledge, lack of long context management over multiple files, and overly critical stance on minor issues. Overall, our results position current AI reviewers as complements to, not substitutes for, human reviewers.
Read the full abstract ▾
All authors Seungone Kim, Dongkeun Yoon, Kiril Gashteovski, Juyoung Suk, Jinheon Baek, Pranjal Aggarwal, Ian Wu, Viktor Zaverkin, Spase Petkoski, Daniel R. Schrider, Ilija Dukovski, Francesco Santini, Biljana Mitreska, Yong Jeong, Kyeongha Kwon, Young Min Sim, Dragana Manasova, Arthur Porto, Biljana Mojsoska, Makoto Takamoto, Marko Shuntov, Ruoqi Liu, Hyunjoo Jenny Lee, Niyazi Ulas Dinç, Yehhyun Jo, Sunkyu Han, Chungwoo Lee, Huishan Li, Esther H. R. Tsai, Ergun Simsek, Khushboo Shafi, Yeonseung Chung, Jihye Park, Aleksandar Shulevski, Henrik Christiansen, Yoosang Son, Elly Knight, Amanda Montoya, Jeongyoun Ahn, Christian Langkammer, Heera Moon, Changwon Yoon, Nikola Stikov, Mooseok Jang, Edward Choi, Junhan Kim, Yeon Sik Jung, Woo Youn Kim, Jae Kyoung Kim, Ishraq Md Anjum, Hyun Uk Kim, Drew Bridges, Carolin Lawrence, Xiang Yue, Alice Oh, Akari Asai, Sean Welleck, Graham Neubig
Machine Learning · Cross-listed from cs.CV
Yuxi Liu, Zekun Zhang, Yixiang Cai et al. (+3 more)
Diffusion Transformers (DiTs) have revolutionized high-fidelity
video generation, yet their $\mathcal{O}(L^2)$
attention complexity poses a formidable bottleneck for long-sequence synthesis. While recent sparse-linear
attention hybrids aim to mitigate this, their performance severely degrades at extreme sparsity due to the "RoPE Dilemma": standard linear
attention fails to preserve the orthogonal relative-position structure of 3D Rotary Position
Embeddings (RoPE), neutralizing vital distance awareness. To address this, we propose RoPeSLR, a 3D RoPE-guided Sparse-LowRank
attention framework. We establish that under empirically validated assumptions, the DiT
attention manifold admits a decoupling into a high-frequency semantic spike set (bounded by $\mathcal{O}(L^{3/2})$ sparsity) and an extreme low-rank ($\mathcal{O}(d_h \log L)$) background continuum. Guided by this structural prior, RoPeSLR eschews standard linear
attention for a head-wise low-rank parameterization equipped with a learnable 3D Absolute Positional
Embedding (PE) injection, seamlessly synthesizing long-range relative distance decay. By guaranteeing sub-quadratic sparsity and sub-linear rank growth, RoPeSLR is exceptionally suited for scaling to ultra-long
video inference. Extensive evaluations validate this scalable superiority: at 90% sparsity, RoPeSLR achieves up to $10\times$ fewer FLOPs on Wan2.1-1.3B and delivers a $2.26\times$ end-to-end inference speedup on the ultra-long 100K+
token sequences of HunyuanVideo-13B, all while maintaining near-lossless generation fidelity (less than 1.3% average VBench degradation).
Read the full abstract ▾
All authors Yuxi Liu, Zekun Zhang, Yixiang Cai, Renjia Deng, Yutong He, Kun Yuan
Machine Learning
REFLECTOR: Internalizing Step-wise Reflection against Indirect Jailbreak
Jiachen Ma, Jiawen Zhang, Xiangtian Li et al. (+3 more)
While Large
Language Models (LLMs) demonstrate remarkable capabilities, they remain susceptible to sophisticated, multi-step jailbreak attacks that circumvent conventional surface-level safety
alignment by exploiting the internal generation process. To address these vulnerabilities, we propose Reflector, a principled two-stage framework that internalizes self-reflection within the generation trajectory. Reflector first leverages teacher-guided generation to produce high-quality reflection data for supervised fine-tuning (SFT), establishing structured reflection patterns. It subsequently uses Reinforcement Learning (RL) with outcome-driven and reward-validity supervision to instill robust, autonomous self-reflection capabilities. Empirical results show that Reflector achieves Defense Success Rates (DSR) exceeding 90% against complex indirect attacks while generalizing robustly across diverse threat scenarios. Notably, the framework enhances both task-specific and general utility, yielding a 5.85% gain on GSM8K alongside improved performance on knowledge-intensive
benchmarks. By internalizing trajectory-level safety, Reflector overcomes the fundamental limitations of surface
alignment without significant computational overhead, offering an efficient and scalable solution for the development of safe and capable LLMs.
Read the full abstract ▾
All authors Jiachen Ma, Jiawen Zhang, Xiangtian Li, Bo Zou, Chaochao Lu, Chao Yang
Machine Learning · Cross-listed from eess.SP
AMAR: Lightweight Attention-Based Multi-User Activity Recognition from Wi-Fi CSI
Amirhossein Mohammadi and Hina Tabassum
Wi-Fi-based human activity recognition (HAR) has emerged as a promising approach for contactless sensing, leveraging channel state information (CSI) collected from wireless transceivers. While existing studies have primarily concentrated on single-user scenarios, real-world deployments often involve multi-user settings where concurrent users' movements induce overlapping CSI patterns that challenge conventional classification methods. To address this limitation, this paper introduces an
attention-based multi-user activity recognition (AMAR) framework that formulates HAR as a set prediction problem. The
transformer-based architecture in AMAR leverages learnable query
embeddings acting as specialized activity detectors, enabling the simultaneous identification of multiple activities from composite CSI representations. Moreover, to address deployment constraints, AMAR is designed in an edge-cloud split architecture form where lightweight convolutional networks on edge devices perform initial feature extraction, followed by residual vector quantization that achieves substantial bandwidth reduction while preserving activity-discriminative information. The cloud component performs final activity prediction through
attention-based set matching, enabling the system to handle varying occupancy levels. Across classroom, meeting-room, and empty-room environments, on average AMAR nearly doubles the rate of perfectly predicting all concurrent activities compared to the best baseline. Moreover, it achieves an $F_1$-score of 53.4% compared to 45.6% for the best
benchmark, and reduces occupancy estimation error by 74%, while minimizing bandwidth substantially.
Read the full abstract ▾
Machine Learning
Design for Manufacturing: A Manufacturability Knowledge-Integrated Reinforcement Learning Framework for Free-Form Pipe Routing in Aeroengines
Caicheng Wang, Zili Wang, Shuyou Zhang et al. (+4 more)
Design for manufacturing plays a critical role in advanced aeroengine development, where complex components necessitate careful consideration of manufacturability. However, current practices in pipe routing remain largely decoupled from down-stream manufacturing, leading to labor-intensive, trial-and-error iterations to achieve manufacturable designs. To address this problem, this study proposes the Frenet-based pipe routing optimization (FPRO) framework, a manufacturability knowledge-integrated reinforcement learning approach for free-form pipe design in aeroengines. FPRO formulates the routing problem as a boundary value problem in the Frenet frame. In this framework, the pipe path is represented by
curvature and torsion profiles, which are generated using cubic Hermite interpolation. To integrate design and manufacturing, domain-specific manufacturing knowledge is embedded as constraints on the permissible ranges of
curvature and torsion. The path optimization is performed using the proximal policy optimization
algorithm with
stochastic exploration and a stage-guided reward mechanism. A unified mapping formulation then translates the optimized path into motion trajectories for the bending die, enabling direct fabrication on a six-axis free-bending machine. Experimental results demonstrate that FPRO consistently generates collision-free, manufacturable paths with smoother
geometric profiles compared to Cartesian-based methods. It also achieves faster convergence and superior performance in terminal
alignment, path length, obstacle avoidance, and manufacturability compared to state-of-the-art reinforcement learning baselines. Real-world validation confirms the close
geometric correspondence between the manufactured pipe and its digital design, validating the practical feasibility of FPRO.
Read the full abstract ▾
All authors Caicheng Wang, Zili Wang, Shuyou Zhang, Yongzhe Xiang, Zheyi Li, Liangyou Li, Jianrong Tan
Machine Learning
AVSD: Adaptive-View Self-Distillation by Balancing Consensus and Teacher-Specific Privileged Signals
Duy Nguyen, Hanqi Xiao, Archiki Prasad et al. (+7 more)
Self-
distillation enables
language models to learn on-policy from their own trajectories by using the same model as both student and teacher, with the teacher being conditioned on privileged information unavailable to the student. Such information can come in different types or views, such as solutions, demonstrations, feedback, or final answers. This setup provides dense
token-level feedback without relying on a separate external model, but creates a fundamental asymmetry: the teacher may rely on view-specific information that the student cannot access at inference time. Moreover, the best type of privileged information is often task-dependent, making it difficult to choose a single teacher view. In this work, we address both these challenges jointly by introducing AVSD (Adaptive-View Self-
Distillation), a novel method of self-
distillation with multiple privileged-information views, which reconstructs
token-level supervision by separating stable cross-view consensus from view-specific residual signals. AVSD identifies the consensus signal shared across views, which provides a reliable update direction, and then selectively adds the view-specific residual signal to adjust the update magnitude when it both aligns with the consensus direction and remains proportionate to the consensus signal. Experiments on math competition
benchmarks (AIME24, AIME25, and HMMT25) show that AVSD consistently outperforms both single-view self-
distillation baselines and GRPO, achieving average Avg@8 gains of 3.1% and 2.2% over the strongest baselines on Qwen3-8B and Qwen3-4B, respectively. Moreover, on code-generation
benchmarks (Codeforces, LiveCodeBench v6) using Qwen3-8B, AVSD outperforms the single-view self-
distillation baseline by 2.4% on average.
Read the full abstract ▾
All authors Duy Nguyen, Hanqi Xiao, Archiki Prasad, Zaid Khan, Anirban Das, Austin Zhang, Sambit Sahu, Hyunji Lee, Elias Stengel-Eskin, Mohit Bansal
Machine Learning
Same Target, Different Basins: Hard vs. Soft Labels for Annotator Distributions
Mirerfan Gheibi and Gashin Ghazizadeh
When annotators disagree, that disagreement can reflect epistemic uncertainty rather than simple label noise. We study hard-label delivery as an alternative to the usual choices of collapsing votes to a single label or training directly on the empirical soft-label distribution. We focus on two primary hard-label methods: multipass, which cycles through observed votes while keeping the
dataset size fixed, and
stochastic label sampling (SLS), which samples one label per example at the start of each epoch. On CIFAR-10H, we find that when only a small number of annotations per example is available, hard-label delivery improves over soft-label training, with larger improvements where the sparse empirical target is farther from the full annotator distribution. When full annotator distributions are available, both hard-label methods match soft-label training. We use deterministic control as an ablation of multipass and shuffled SLS as a control that breaks the example-to-distribution match. We also show that SLS and soft-label cross-
entropy optimize the same expected objective. Hard-label delivery also converges to flatter basins, with supporting descriptive evidence from OOD detection on SVHN and CIFAR-100. Overall, these results suggest that multipass is a strong practical default when raw vote counts are available, while SLS offers a lightweight alternative that remains competitive when only a few votes per example are available and matches soft-label training when full annotator distributions are available.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.CR
Trusted Weights, Treacherous Optimizations? Optimization-Triggered Backdoor Attacks on LLMs
Yifei Wang, Yida Yang, Tianlin Li et al. (+3 more)
Inference optimization aims to minimize the latency and resource consumption of LLM inference while preserving output quality, making large-scale deployment practical and cost-effective. However, optimized execution can introduce small numerical inconsistencies from the original model. We reveal that this inconsistency not only causes the model's outputs to diverge, but more critically can introduce hidden
backdoors. The
backdoor remains dormant under standard unoptimized execution and is activated only when inference optimization is enabled, allowing it to evade existing
backdoor detection pipelines. We first introduce the Input-Specific Optimization
Backdoor (ISOB) to demonstrate that optimization-induced differences can cause wrong predictions. However, ISOB remains input-specific and cannot establish a universal optimization-triggered
backdoor. To overcome this limitation, we design the Universal Optimization
Backdoor (UOB). The backdoored model stays benign under unoptimized execution but activates when inference optimization is enabled. We conduct extensive experiments across seven mainstream open-source LLMs, four tasks, and three optimization backends. UOB reaches up to 100% attack success while largely preserving clean accuracy. To mitigate this vulnerability, we design three defense methods that reduce the
backdoor ASR to at most 0.02 while preserving clean accuracy. These results reveal inference optimization as a new LLM security attack surface and motivate defenses against test-deployment disagreement.
Read the full abstract ▾
All authors Yifei Wang, Yida Yang, Tianlin Li, Xiaohan Zhang, Xiaoyu Zhang, Li Pan
Machine Learning · Cross-listed from math.OC
Time-Dependent PDE-Constrained Optimization via Weak-Form Latent Dynamics
April Tran, Terry Haut, David Bortz et al. (+1 more)
Optimization problems constrained by high-dimensional, time-dependent partial differential equations require repeated forward and sensitivity solves, making high-fidelity optimization computationally prohibitive in many-query design and control settings. We present a weak-form latent-space reduced-order modeling framework for accelerating gradient-based PDE-constrained optimization. The proposed approach builds on Weak-form Latent Space Dynamics Identification (WLaSDI), which compresses high-dimensional solution trajectories into a low-dimensional latent representation and identifies parametric latent dynamics using weak-form system identification. By avoiding explicit numerical differentiation of training trajectories, the weak-form improves robustness to noisy data and yields more reliable surrogate dynamics for optimization. We formulate the resulting reduced PDE-constrained optimization problem and derive both direct-sensitivity and adjoint-based gradient expressions for the learned latent dynamics, enabling scalable gradient evaluation with respect to design parameters. The framework is demonstrated on three time-dependent
benchmark problems: thermal radiative transfer for optimal hohlraum design, the two-stream instability Vlasov-Poisson system, and the inviscid Burgers equation. Across these examples, WLaSDI produces accurate optimal designs, remains robust under noisy training data, and delivers substantial computational savings, including speedups of up to five orders of magnitude relative to full-order optimization. These results demonstrate that weak-form latent dynamics provide an efficient and noise-robust surrogate foundation for gradient-based optimization of complex time-dependent PDE systems.
Read the full abstract ▾
All authors April Tran, Terry Haut, David Bortz, Youngsoo Choi
Machine Learning
The General Theory of Localization Methods
Congwei Song
This paper proposes a general machine learning framework called the localization method, which is fundamentally built on two core concepts: localization kernels and local means – key components that underpin the self-
attention mechanism. To establish a rigorous theoretical foundation, the framework is formally defined through two essential pillars: the formulation of the local(-ized) model and the localization trick. We systematically investigate the connections between the localization method and a wide range of existing machine learning models/methods, including (but not limited to) kernel methods, lazy learning, the MeanShift
algorithm, relaxation labeling, Hopfield networks, local linear
embedding (LLE), fuzzy inference, and denoising autoencoders (DAEs). By dissecting these relationships, we clarify the broader theoretical significance of the localization method and demonstrate its practical applicability across diverse machine learning tasks. Furthermore, we explore advanced extensions of the framework, such as adaptive kernels, hierarchical local models, and non-local models. Notably, we show that the
Transformer – a cornerstone of modern sequence modeling – can be constructed using hierarchical local models, revealing the ability of the localization method to unify and generalize state-of-the-art architectures. This work not only provides a unified theoretical lens to reinterpret existing models but also offers new methodological tools for designing flexible, data-adaptive learning systems.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.CV
Accelerating Video Inverse Problem Solvers with Autoregressive Diffusion Models
Taesung Kwon, Jonghyun Park, Hyungjin Chung et al. (+1 more)
Diffusion models provide powerful priors for zero-shot
video inverse problems, but their real-time deployment is hindered by two inefficiencies: high initial latency caused by holistic
video restoration, and low throughput resulting from multiple VAE passes to enforce measurement consistency in pixel space. To overcome these limitations, we propose Autoregressive
Video Inverse problem Solver (AVIS). The AVIS framework leverages autoregressive
video diffusion models to restore
videos in a streaming manner, naturally eliminating latency bottlenecks. Specifically, AVIS initializes reverse
diffusion with a measurement-consistent estimate, reducing the required sampling steps. Compared to leading non-autoregressive solvers, AVIS drastically reduces initial latency from 114s to 4s and increases throughput from 0.71 to 1.18 FPS while achieving superior restoration quality. We further introduce a highly accelerated variant, dubbed AVIS Flash, that enforces measurement consistency solely on the first chunk. AVIS Flash substantially boosts throughput to 5.91 FPS on a single RTX 4090 GPU while maintaining competitive performance and achieving a favorable efficiency-performance trade-off, paving the way toward real-time deployment.
Read the full abstract ▾
All authors Taesung Kwon, Jonghyun Park, Hyungjin Chung, Jong Chul Ye
Machine Learning
Dynamic Shapley Computation
Xuan Yang, Hsi-Wen Chen, Ming-Syan Chen et al. (+1 more)
Shapley-based data valuation provides a principled way to quantify the contribution of training data, but its high computational cost makes it impractical in dynamic settings where tasks and training players evolve. Existing methods treat Shapley computation as a one-shot process and collapse contributions into aggregated scores, preventing reuse and requiring recomputation under any change. We introduce a new perspective that represents Shapley values as a player-by-task
matrix and formulates dynamic valuation as a structured
matrix maintenance problem. We exploit the fact that each task depends on a small subset of training players and that similar tasks yield similar valuations, leading to utility locality and coalition locality. Based on these insights, we propose D-Shap, a dynamic valuation framework that enables efficient updates by modifying only a small portion of the
matrix: new task valuations are inferred via structure-aware interpolation, while updates induced by new players are confined to affected local
matrix blocks. To eliminate the need for pre-specified evaluation tasks, we introduce self-valuation, which constructs the initial
matrix directly from training data, supported by scalable subset reuse and coverage-aware anchor selection. Experiments across diverse models show that D-Shap performs task updates in milliseconds and reduces the cost of player updates by up to three orders of magnitude, while achieving valuation quality competitive with full recomputation.
Read the full abstract ▾
All authors Xuan Yang, Hsi-Wen Chen, Ming-Syan Chen, Jian Pei
Machine Learning
SURF: Steering the Scalarization Weight to Uniformly Traverse the Pareto Front
Liuyuan Jiang, Chentong Huang and Lisha Chen
Scalarization is widely used in multi-objective optimization owing to its simplicity and scalability. In many applications, the goal is to generate solutions that represent diverse user preferences, ideally with uniform coverage of the Pareto front (PF). However, uniformly sampling scalarization weights usually induces non-uniform coverage of the PF. We explain this mismatch through the geometry of the scalarization path. As the weight varies, the corresponding solutions traverse the PF at a generally nonuniform speed. Integrating this speed gives cumulative arc length. After normalization by the total length, it defines a cumulative distribution function (CDF) on weights whose inverse selects points at equal arc-length intervals. Building on this insight, we propose SURF (Sampling Uniformly along the Pareto Front). For structured problems, including bi-objective bandits, we derive analytical expressions for this CDF map and the resulting PF-aware weight sampling rule. For general problems, SURF alternates between CDF reconstruction and weight sampling. Theoretically, we show that under provable conditions for bi-objective LS, SURF's CDF reconstruction error converges linearly to an $O(N^{-2})$ discretization floor. Empirically, experiments on bandits, multi-objective-gymnasium, and multi-objective LLM
alignment demonstrate that SURF efficiently achieves more uniform PF coverage than baselines.
Read the full abstract ▾
Machine Learning
Matryoshka Concept Bottleneck Models
Ziye Chen, Hongbin Lin, Jie Li et al. (+1 more)
Concept Bottleneck Models (CBMs) have emerged as a prominent paradigm for interpretable deep learning, learning by grounding predictions in human-understandable concepts. However, their practical deployment is hindered by the high cost of test-time intervention, as correcting model errors typically requires human experts to manually inspect and verify a large set of predicted concepts. Existing approaches suffer from a fundamental structural limitation: they either adopt a single static concept set, forcing experts to exhaustively annotate concepts and incurring prohibitive intervention costs, or train multiple models tailored to different concept budgets, resulting in substantial computational and maintenance overhead. To address this challenge, we propose the Matryoshka Concept Bottleneck Model (MCBM), a unified architecture that enables adaptive concept utilization within a single model. Inspired by Matryoshka Representation Learning, MCBM organizes concepts into a nested hierarchy based on maximum relevance and minimum redundancy, allowing inference at multiple levels of conceptual granularity without retraining. Theoretically, we show that MCBM reduces the expected intervention costs from linear to logarithmic order, $O(\log K)$, while guaranteeing monotonic performance improvement. Empirically, extensive experiments demonstrate that MCBM matches the performance of independently trained models while enabling dynamic and efficient expert interaction.
Read the full abstract ▾
All authors Ziye Chen, Hongbin Lin, Jie Li, Lijie Hu
Machine Learning
Compositional Transduction with Latent Analogies for Offline Goal-Conditioned Reinforcement Learning
Junseok Kim, Dohyeong Kim, Mineui Hong et al. (+1 more)
Compositional generalization is essential for reaching unseen goals under novel contextual variations in offline goal-conditioned reinforcement learning (GCRL), where a generalist goal-reaching
agent must be learned from limited data. Most prior approaches pursue this via trajectory stitching over temporally contiguous segments, which limits composing behaviors across varying contexts. To overcome this limitation, we formalize analogy transduction as synthesizing new plans by composing task-endogenous analogies with given contexts and propose a novel analogy representation tailored for it. Grounded in our theory, this analogy representation captures what changes under optimal task execution, remains invariant to contextual variations, and is sufficient for optimal goal reaching. We further contend that generalization to unseen analogy-context pairs is a practical obstacle in analogy transduction, and introduce a new approach for offline GCRL that enables analogy transduction beyond seen pairs to unseen combinations. We empirically demonstrate the effectiveness of our approach on OGBench manipulation environments, substantially outperforming prior methods that do not perform analogy transduction. Project page: https://rllab-snu.github.io/projects/CTA/
Read the full abstract ▾
All authors Junseok Kim, Dohyeong Kim, Mineui Hong, Songhwai Oh
Machine Learning
Mechanistic Interpretability for Learning Assurance of a Vision-Based Landing System
Romeo Valentin, Olivia Beyer Bruvik, Marc R. Schlichting et al. (+1 more)
EASA's learning-assurance guidance requires data-driven aviation systems to build and monitor their own situation representation, yet for neural networks the technical means to provide such evidence remain an open problem. We address this gap for a
vision-based aircraft landing system: we propose that a minimally assurable model must at least be shown to separate content from style in its own situation representation. Showing that the model's predictions then rely largely on the contentful representation components leads to a concrete assurance path. To demonstrate this assurance path on a concrete model we train a
vision transformer model for runway keypoint regression on the LARDv2
dataset. The model, which acts as the subject for our assurance demonstration, produces per-patch
embeddings that we decompose into interpretable atoms via K-SVD sparse dictionary learning. A qualitative visualization confirms that contentful atoms track task-relevant runway structure and stylistic atoms track domain-specific appearance, and the regression head is shown to place almost all of its linear weight on contentful atoms. We further build on the content/style separation and define out-of-model-scope (OOMS) detection, a novel runtime assurance approach directly monitoring the model's situation representation. OOMS monitoring is complementary to operational design domain and output-space out-of-distribution monitoring and addresses concrete requirements of the recent EASA guidance. By directly analyzing a model's situation representation both at test time and runtime, this work delivers the first concrete piece of the representation-level evidence that EASA learning-assurance guidance demands, and points to mechanistic interpretability as a practical building block of future aviation safety cases.
Read the full abstract ▾
All authors Romeo Valentin, Olivia Beyer Bruvik, Marc R. Schlichting, Mykel J. Kochenderfer
Machine Learning · Cross-listed from cs.CL
Self-Training Doesn't Flatten Language – It Restructures It: Surface Markers Amplify While Deep Syntax Dies
Ming Liu
Successive self-training on a
language model's own outputs is widely characterized as a process of flattening: diversity drops, distributions narrow, and the text becomes "more like itself." We provide evidence that this characterization is incomplete. Across eleven generations of self-training on five models (GPT-2 124M, Pythia-410M, Pythia-1.4B, OPT-1.3B, Pythia-2.8B),
language is not flattened uniformly – it is restructured. Surface markers (discourse connectives, hedges, em-dashes) rise, while mid- and deep-syntactic structures (questions, parentheticals, passives, subjunctives) collapse. We formalize this asymmetric collapse as the Structural Depth Hypothesis (SDH): the per-generation decay rate of a linguistic feature is predicted primarily by its structural depth – the number of nested syntactic dependencies it requires – and only secondarily by its generation-zero output frequency. Pooling 17-feature panels from five models spanning three architecture families (N=85), the pooled Spearman correlation is rho=0.540 (p < 10^{-6}; cluster-bootstrap 95% CI [0.434, 0.634]), while frequency is a substantially weaker predictor (rho=0.225). A matched human-text fine-tuning control yields rho=0.039 (p=0.88), confirming the gradient is self-training-specific. We further document a Superficial Complexity Paradox: aggregate complexity proxies (dep-tree depth, TTR, word length) all rise as the underlying clause structure dies, with direct implications for training-data curation and LLM-text detection.
Read the full abstract ▾
Machine Learning
Unsupervised clustering and classification of upper limb EMG signals during functional movements: a data-driven
L. F. Salazar Álvarez, D. Escobar-Saltarén, M. B. Salazar Sánchez et al. (+1 more)
This study presents a comprehensive approach for the clustering and classification of upper-limb surface electromyography (sEMG) signals during functional reach and grasp movements. The methodology was applied to the NINAPRO DB4
dataset, which provides multichannel EMG recordings of 52 gestures. A four-stage pipeline was designed, including signal preprocessing, fea-ture extraction, gesture selection via hierarchical clustering, and comparative model evaluation. Preprocessing involved a fourth-order low-pass filter (0.6 Hz) and
Hilbert envelope transformation, effectively reducing noise and enhancing signal clarity. Feature extraction yielded 26 temporal and frequency-domain met-rics, which were later refined using visual analysis, mutual information, principal component analysis, and decision tree importance scores. A final subset of five key features was selected for classification tasks. Gesture selection was per-formed through hierarchical clustering using Mahalanobis distance, resulting in six representative movements that balanced biomechanical diversity and compu-tational efficiency. A 200 ms window was identified as optimal for temporal seg-mentation based on stability and physiological plausibility. Classifier models were evaluated in two stages. Automated comparison using PyCaret identified Extra Trees (ET) and Artificial Neural Networks (ANN) as top performers. Sub-sequent independent training confirmed their stability and generalization capac-ity, with ANN showing progressive learning and ET maintaining robust, con-sistent results. The findings support the implementation of adaptive, low-latency control strategies for myoelectric prostheses and provide a scalable pipeline for future real-time applications.
Read the full abstract ▾
All authors L. F. Salazar Álvarez, D. Escobar-Saltarén, M. B. Salazar Sánchez, S. C. Henao-Aguirre
Machine Learning
ReversedQ: Opportunities for Faster Q-Learning in Episodic Online Reinforcement Learning
Sofia R. Miskala-Dinc and Aviva Prins
We study model-free Q-learning in finite-horizon episodic Markov Decision Processes (MDPs) with stationary dynamics across episodes. We identify a central issue in nascent model-free posterior-sampling works: the reliance on delayed learning in order to prove theoretical guarantees. In particular, we identify three opportunities for faster learning - (i) value-function update order, (ii) update frequencies, and (iii) value-function initialization. Using Wang et al.'s RandomizedQ as a basis, we illustrate these changes and their individual (as well as cumulative) impact in multiple empirical studies. We find that our combined modifications, termed ReversedQ, improve scaled mean cumulative reward compared to RandomizedQ, from 9.53% to 78.78% in the Bidirectional Diabolical Combination Lock (BDCL), and from 21.76% to 61.81% in a chain MDP.
Read the full abstract ▾
Machine Learning
TriForces: Augmenting Atomistic GNNs for Transferable Representations
Ali Ramlaoui, Alexandre Duval, Hannah Bull et al. (+4 more)
Machine learning interatomic potentials (MLIPs) achieve excellent accuracy when trained on large Density Functional Theory (DFT) data. To be useful in practice, they must often be adapted to target chemistries using small and expensive task-specific
datasets. However, MLIPs transfer inconsistently across domains, with representations that often loose accessible composition and structure information. To address this, we present TriForces, a model-agnostic three-stream framework that separates composition and structure information, combined with self-supervised learning to preserve transferable representations. TriForces improves performance on MatBench and QM9 over baselines without needing DFT labels and enables efficient similar structure
retrieval through its learned latent space. On OMat24, in limited-data training regime, TriForces reduces energy MAE by 57% at 20K samples only and improves force MAE across sample sizes. We release pretrained TriForces variants across multiple MLIP architectures with code at https://github.com/Ramlaoui/triforces.
Read the full abstract ▾
All authors Ali Ramlaoui, Alexandre Duval, Hannah Bull, Victor Schmidt, Hugues Talbot, Fragkiskos D. Malliaros, Joseph Musielewicz
Machine Learning
Deep Learning Surrogates for Emulating Stochastic Climate Tipping Dynamics
Adeline Hillier, Jennifer Sleeman, Jay Brett et al. (+3 more)
This work explores a dynamics-informed Temporal Fusion
Transformer (TFT) as a data-driven surrogate for computationally intensive Earth system simulations. Focusing on multivariate time series describing global ocean transport, we demonstrate the surrogate's ability to forecast tip events across thousands of time steps. The data involve up to 21 non-stationary time series in addition to static
covariates describing free parameters and initial conditions. Modifications to the architecture and objective function yield a surrogate that anticipates the timing of Atlantic and Pacific collapses to high fidelity and captures the
stochastic uncertainty in transition timing across ensemble predictions. The learned surrogate achieves a 465x computational speedup over the numerical simulator while maintaining differentiability with respect to parameters and initial conditions.
Read the full abstract ▾
All authors Adeline Hillier, Jennifer Sleeman, Jay Brett, Caroline Tang, Jenelle Millison, Anand Gnanadesikan
Machine Learning · Cross-listed from cs.AI
Mahjax: A GPU-Accelerated Mahjong Simulator for Reinforcement Learning in JAX
Soichiro Nishimori, Shinri Okano, Keigo Habara et al. (+3 more)
Riichi Mahjong is a multi-player, imperfect-information game characterized by stochasticity and high-dimensional state spaces. These attributes present a unique combination of challenges that mirror complex real-world decision-making problems in reinforcement learning. While prior research has heavily relied on supervised learning from human play logs to pre-train the policy,
algorithms capable of learning tabula rasa (from scratch) offer greater potential for general applicability, as evidenced by the AlphaZero lineage. To facilitate such research, we introduce Mahjax, a fully vectorized Riichi Mahjong environment implemented in JAX to enable large-scale rollout parallelization on Graphics Processing Units (GPUs). We also provide a high-quality visualization tool to streamline debugging and interaction with trained
agents. Experimental results demonstrate that Mahjax achieves throughputs of up to 2 million and 1 million steps per second on eight NVIDIA A100 GPUs under the no-red and red rules, respectively. Furthermore, we validate the environment's utility for reinforcement learning by showing that
agents can be trained effectively to improve their rank against baseline policies.
Read the full abstract ▾
All authors Soichiro Nishimori, Shinri Okano, Keigo Habara, Sotetsu Koyamada, Eason Yu, Masashi Sugiyama