Machine Learning · Cross-listed from cs.RO
Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models
Yifu Yuan, Yaoting Huang, Xianze Yao et al. (+20 more)
We introduce
Embodied-R1.5, a unified
Embodied Foundation Model (EFM) that integrates comprehensive
embodied reasoning capabilities, spanning
embodied cognition, task planning, correction, and pointing, within a single architecture toward general physical intelligence. Leveraging three automated data construction pipelines to significantly expand the data coverage of critical capabilities, we build a large-scale data system of over 15B
tokens, and design a multi-task balanced RL recipe to alleviate heterogeneous task conflicts. We further introduce a Planner-Grounder-Corrector (PGC) closed-loop framework that enables a single model to autonomously execute and self-correct over long-horizon tasks. With only 8B parameters,
Embodied-R1.5 achieves SOTA on 16 out of 24
embodied VLM
benchmarks, surpassing leading models like Gemini-Robotics-ER-1.5 and GPT-5.4. Benefiting from the internalized
embodied capabilities,
Embodied-R1.5 can be fine-tuned into a VLA with only a small amount of data, outperforming leading VLA models like $\pi_{0.5}$ across 4 popular manipulation
benchmark suites. We further conduct extensive zero-shot real-
robot experiments, validating performance in instruction following, affordance grounding, articulated object manipulation, and long-horizon complex tasks, demonstrating strong generalization to the physical world. We open-source model weights,
datasets, training code, and EmbodiedEvalKit, an evaluation framework tailored for
embodied tasks, to facilitate future research in EFMs.
Read the full abstract ▾
All authors Yifu Yuan, Yaoting Huang, Xianze Yao, Yutong Li, Shuoheng Zhang, Linqi Han, Pengyi Li, Jiangeng Sun, Wenting Jia, Zhao Zhang, Yuhao Liu, Ruihao Liao, Yucheng Hu, Qiyu Wu, Yuxiao Li, Zibin Dong, Fei Ni, Yan Zheng, Shuyang Gu, Yi Ma, Hongyao Tang, Han Hu, Jianye Hao
Machine Learning
Neural networks become centroid classifiers when trained on heavily corrupted inputs
Justin Tahmassebpur, Asadullah Bhuiyan, Hyejin Kim et al. (+1 more)
Learning from imperfect data is a central theme in machine learning, connecting practical questions of robustness to fundamental questions of learnability. Here we examine attribute noise: learning from corrupted inputs while keeping the labels intact, a setting that has received considerably less analytical
attention than its label-noise counterpart. We consider two types of corruption models: additive noise and replacement noise. Through experiments with multi-layer perceptrons (MLPs) on corrupted classification
datasets, we find that neural networks remain robust, maintaining well-above-chance accuracy even when inputs are >90% corrupted – far beyond human recognition. To understand this robustness, we analyze infinite-width networks in the heavy-corruption regime using a mean-field-inspired approach and derive a leading-order decision rule for the classification outcome: the network implements a prototype rule, the nearest-class-mean, assigning each test point to the class whose training-set average it most closely resembles. This leading-order decision rule is universal across a broad range of MLP architectures, holding for any depth, as well as a wide class of activation functions and noise distributions. The same centroid mechanism closely matches finite-width network behavior in our experiments and provides an interpretable and analytically tractable account of why learning can succeed even when individual training examples carry almost no signal.
Read the full abstract ▾
All authors Justin Tahmassebpur, Asadullah Bhuiyan, Hyejin Kim, Omri Lesser
Machine Learning · Cross-listed from physics.ins-det
SPADE: Split-and-Delay Embeddings for Autoregressive High-Granularity Calorimeter Simulation
Joschka Birk, Frank Gaede, Anna Hallin et al. (+3 more)
We introduce SPADE (SPlit And Delay
Embeddings), an autoregressive
transformer for sequences whose
tokens carry multiple features. Rather than
embedding these features jointly, SPADE embeds them independently. Delaying each feature stream relative to the previous one allows intra-
token correlations to be learned by the standard self-
attention mechanism. Applied to point-cloud calorimeter shower generation in the highly granular ILD detector, SPADE is competitive with the state of the art AllShowers model on
photon showers, and substantially outperforms its VQ-VAE-based predecessor OmniJet-$\alpha_C$. The mechanism is applicable to any generative task with multi-feature
tokens, enabling LLM-style pretraining workflows for higher-dimensional data.
Read the full abstract ▾
All authors Joschka Birk, Frank Gaede, Anna Hallin, Gregor Kasieczka, Martina Mozzanica, Henning Rose
Machine Learning · Cross-listed from astro-ph.CO
Interpretable Neural Marked Statistics for Cosmological Inference
Federico Semenzato, Benjamin D. Wandelt, Michele Liguori et al. (+1 more)
Recovering cosmological information beyond the power spectrum is a central goal for upcoming cosmological surveys, since late-time non-Gaussian signal in the matter density cannot be accessed through two-point statistics alone. Marked statistics fold part of this information back into the two-point level by reweighting the field with non-linear functions. We propose a neural marking scheme to generalize this process through a set of interpretable, physically motivated transformations that directly allow to interpret the gain in cosmological information at the morphological level. We employ a contrastive learning objective to align learnable marked summaries with the underlying cosmological parameters. At $k_{\max}=0.2\,h\mathrm{Mpc}^{-1}$, our neural mark tightens the marginalized constraint on $\sigma_8$ by $2.9\times$ and on $\Omega_m$ by $1.8\times$ compared to classical marks, breaking the $\Omega_m-\sigma_8$ degeneracy at the Fisher information level. It further reduces the parameter MSE across our cosmological parameter prior by $1.45\times$ over the best classical mark. The learned latent geometry aligns with the $\Omega_m$ and $\sigma_8$ directions in parameter space, indicating that the contrastive objective recovers the dominant axes of cosmological information. Our approach opens the door to more powerful, interpretable summary statistics for cosmological inference.
Read the full abstract ▾
All authors Federico Semenzato, Benjamin D. Wandelt, Michele Liguori, Alvise Raccanelli
Machine Learning
FlowBank: Query-Adaptive Agentic Workflows Optimization through Precompute-and-Reuse
Lingzhi Yuan, Chenghao Deng, Fangxu Yu et al. (+3 more)
Large
Language Model (LLM)-based multi-
agent systems are increasingly powerful, but current agentic workflow optimization paradigms make an unsatisfying trade-off. Task-level methods spend substantial offline compute yet deploy only a single workflow, leaving complementary candidates unused, while query-level methods synthesize a new workflow per query at substantial inference cost. Our motivating analysis shows these paradigms are more complementary than competing: workflows discovered during offline search often solve different subsets of queries, and many queries handled by expensive query-level generation can already be solved by cheaper precomputed workflows. This suggests a different objective: rather than searching for one universally best workflow or regenerating one per instance, we should build a compact bank of reusable, complementary workflows and select among them adaptively at inference time. Doing so requires solving three coupled problems: generating complementary rather than redundant candidates, compressing them into a small deployable portfolio, and assigning each query to the right workflow under a performance-cost trade-off. To this end, we present FlowBank, a three-stage framework for portfolio-based agentic workflow optimization. Diversifying proposes DiverseFlow to steer search toward under-covered queries and produce a high-coverage candidate pool. Curating proposes CuraFlow to compress this pool into a compact portfolio with minimal redundancy. Matching casts deployment as edge-value prediction on a query-workflow bipartite graph and routes each incoming query to the portfolio member with the best predicted utility. Across five
benchmarks, FlowBank achieves the highest average score among the evaluated methods while remaining cost-competitive, improving over the strongest automated and handcrafted baselines by 4.26% and 14.92% relative, respectively.
Read the full abstract ▾
All authors Lingzhi Yuan, Chenghao Deng, Fangxu Yu, Souradip Chakraborty, Mohammad Rostami, Furong Huang
Machine Learning
FreeBridge: Variational Schrödinger Bridges for Cellular Transition Dynamics
Xurui Wang, Qin Ren, Jun Ma et al. (+2 more)
High-content imaging assays quantify cellular responses to chemical and genetic perturbations, yet continuous trajectories of individual
cells are unobservable because
cells are chemically fixed at acquisition. Perturbation modeling therefore reduces to inferring
stochastic transport between control and treated populations observed only as separate marginals. While recent generative models achieve strong end-point
alignment, boundary consistency does not determine intermediate
evolution: multiple
stochastic processes may connect identical marginals while traversing regions unsupported by observed single-
cell morphologies. We introduce FreeBridge, a Schrödinger Bridge formulation for single-
cell transition modeling under endpoint-only supervision. FreeBridge defines atomic states as instance-segmented single-
cell representations, establishing a fixed cellular
manifold, and learns
stochastic transport constrained within this geometry via empirical latent support regularization. Across BBBC021, RxRx1, and JUMP, FreeBridge maintains competitive or improved endpoint fidelity and mechanism-of-action retention under a unified evaluation protocol; on BBBC021, it further reduces intermediate support violations. These findings highlight the importance of
geometric grounding for biologically interpretable perturbation dynamics. Project page: https://y-research-sbu.github.io/FreeBridge/.
Read the full abstract ▾
All authors Xurui Wang, Qin Ren, Jun Ma, Haibin Ling, Chenyu You
Machine Learning · Cross-listed from cs.MA
Phi-Actor-Critic: Steering General-Sum Games to Pareto-Efficient Correlated Equilibria
Wongyu Lee, Francesco Lelli, Omran Ayoub et al. (+1 more)
Real-world multi-
agent systems, from traffic coordination to resource allocation, are often modeled as general-sum games where individual incentives conflict with collective welfare. In these settings, the central challenge is not merely finding an
equilibrium, but selecting socially desirable outcomes among many suboptimal Nash
equilibria. Standard deep multi-
agent reinforcement learning (MARL) methods struggle with this problem, as value-decomposition approaches are constrained by monotonicity assumptions and policy-gradient methods often converge to stable but socially inefficient
equilibria. To address this limitation, we propose $\Phi$-Actor-Critic ($\Phi$-AC), a framework that leverages swap regret minimization to steer learning toward high-welfare correlated
equilibria (CE). To make counterfactual regret estimation tractable in deep MARL, $\Phi$-AC employs a centralized
attention critic that predicts vector-valued regrets in a single forward pass, avoiding computationally expensive counterfactual simulations. We further introduce a Lagrangian-based
equilibrium selection mechanism that optimizes social welfare while enforcing stability through regret constraints. Experiments on
matrix games, Multi-
Agent Particle Environments (MPE), and the Melting Pot Harvest scenario demonstrate that $\Phi$-AC learns efficient and stable coordination strategies across diverse mixed-motive settings while maintaining high collective return and competitive
fairness.
Read the full abstract ▾
All authors Wongyu Lee, Francesco Lelli, Omran Ayoub, Massimo Tornatore
Machine Learning · Cross-listed from cs.DS
Fixed-Parameter Tractability of Private Synthetic Data Generation
Badih Ghazi, Cristóbal Guzmán, Pritish Kamath et al. (+3 more)
We study the problem of generating synthetic data under differential
privacy. We establish fixed-parameter tractability (FPT) for this problem where the parameter is the treewidth of the query family's incidence graph. Our
algorithms attain optimal error rates across all regimes and are realized by two different approaches: the first is based on linear programming (LP) and the FPT of the separation problem for the LP dual; the second is based on a subsampled private multiplicative weights method, where we obtain FPT for sampling from Gibbs distributions. Both approaches are unified by a dynamic programming framework over a tree decomposition.
Read the full abstract ▾
All authors Badih Ghazi, Cristóbal Guzmán, Pritish Kamath, Alexander Knop, Ravi Kumar, Pasin Manurangsi
Machine Learning · Cross-listed from eess.AS
Massive Open-Vocabulary Keyword Spotting
Leonor Barreiros, Raul Monteiro, Afonso Mendes et al. (+1 more)
Automatic
speech recognition systems have been shown to under-perform when it comes to transcribing words rarely seen in the training data, namely specialized terminology. Open-vocabulary keyword spotting, combined with contextual biasing, has been shown to mitigate this issue. However, existing systems can only handle glossaries of a few hundred terms without becoming an infeasible bottleneck. We propose a system that stores features with a
memory footprint up to 128 times smaller than a comparable baseline and allows users to process massive databases while remaining open-vocabulary. Without fine-tuning the
speech recognition model, our system achieves a comparable entity recall as uncompressed solutions, even in
languages not seen during training.
Read the full abstract ▾
All authors Leonor Barreiros, Raul Monteiro, Afonso Mendes, Gonçalo M. Correia
Machine Learning
Least-Action-Guided Diffusion for Physical Extrapolation
Zhongxin Yang, Yuanwei Bin, Xiang I.A. Yang et al. (+1 more)
Reliable extrapolation remains a central challenge for generative models in computational physics, because models trained over finite ranges of time, parameters, or geometries may produce physically inconsistent predictions outside the training distribution. We introduce a least-action-principle-guided
diffusion, LAPG, a framework that promotes physical consistency during inference rather than relying solely on constraints imposed during training. The method combines a conditional score-based
diffusion model with an action-derived physical guidance score. In the first stage, the learned score model generates an in-distribution proposal; in the second, an action-based variational prior refines this proposal toward the target out-of-distribution condition. This formulation turns the principle of least action into a differentiable inference-time correction mechanism and provides an alternative to pointwise residual penalties that often require empirical loss balancing. We evaluate LAPG on representative ordinary- and partial-differential-equation systems, including free fall, conservative and dissipative spring-mass dynamics, interacting point
vortices, and potential flow over parameterized airfoils. In temporal, parameter, and
geometric extrapolation tests, LAPG reduces phase drift, preserves dissipative decay, captures
vortex motion, and improves the lift response of airfoil flows compared with training-time physics-informed baselines.
Read the full abstract ▾
All authors Zhongxin Yang, Yuanwei Bin, Xiang I.A. Yang, Shiyi Chen
Machine Learning
RoVE: Rotary Value Embeddings Attention for Relative Position-dependent Value Pathways
Alejandro García-Castellanos, Maurice Weiler and Erik J Bekkers
Rotary Position
Embeddings (RoPE) make
attention scores position-relative but leave the value pathway position-blind: the message sent by a value
token is the same regardless of its distance from the query. We propose RoVE, a parameter-free modification that makes values position-sensitive by rotating them simultaneously with keys, and show that it turns RoPE
attention into attentive convolution. This new perspective unifies several independent formulations of the same operation across computer
vision, robotics, and modern LLM architectures. Trained 124M and 354M GPT-2 models show consistent empirical gains over RoPE on few-shot in-context learning, out-of-distribution perplexity, and long-context
retrieval, with the clearest improvements on tasks that require long-range aggregation.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.MA
Multi-agent rendezvous in fluid flows via reinforcement learning
Bocheng Li, Jingran Qiu and Lihao Zhao
Rendezvous is a critical task for multi-
agent systems, requiring
agents to coordinate to meet at an unspecified location. However, achieving this in fluid environments presents a challenge, as it remains unclear how
agents can exploit underlying fluid kinematics to facilitate convergence. In this study, we adopt a multi-
agent reinforcement learning (MARL) approach to develop physics-informed rendezvous strategies in vortical flows. Compared to a naive strategy, where
agents navigate toward their counterparts, MARL strategies significantly improve the rendezvous rate. MARL strategies also show transferability across varying
vortex intensities,
vortex scales, and swarm sizes. By breaking the
symmetry of the state-action map, MARL strategy leverages a non-intuitive mechanism that prevents
agents from becoming trapped in separate
vortices, thereby enhancing rendezvous success. Additionally, a heuristic strategy is extracted from the learned strategy and also outperforms the naive strategy. Furthermore, a theoretical analysis demonstrates that fluid deformation impedes the rendezvous process. Large finite-time Lyapunov exponents identify where fluid effects separate adjacent
agents, suggesting that targets should be planned in weak-deformation regions. Our findings reveal the important role that
agent-fluid interactions play in multi-
agent tasks and highlight the MARL capability to explore swarm intelligence in complex flow environments.
Read the full abstract ▾
Machine Learning
Federated continual learning: A comprehensive survey on lifelong and privacy-preserving learning over distributed and non-stationary data
Masoume Gholizade, Fabrizio Ruffini, Pietro Ducange et al. (+1 more)
Federated Learning (FL) enables collaborative and
privacy-preserving model training across distributed clients, but most existing FL systems implicitly assume data stationarity. In real-world settings-such as healthcare, industrial IoT (IIOT), cybersecurity, and smart cities-data streams are inherently non-stationary, leading classical FL methods to suffer from performance degradation, instability, and catastrophic forgetting. Continual Learning (CL) addresses learning under evolving data distributions but has been largely studied in centralized settings, overlooking key constraints of
federated systems, including
privacy, limited communication, and client heterogeneity.
Federated Continual Learning (FCL) emerges at the intersection of FL and CL, aiming to support lifelong, adaptive, and
privacy-aware learning over distributed and non-stationary data. This survey provides a comprehensive and systematic overview of FCL. We first present a formal definition of the FCL problem and clarify its distinctive characteristics. We then analyze the limitations of classical FL under non-stationary conditions, highlighting how CL principles support long-term adaptation. To organize the rapidly growing literature, we propose a multi-dimensional taxonomy of FCL approaches. Furthermore, we review representative application domains and data modalities, summarize commonly used evaluation metrics, and discuss experimental perspectives for assessing long-term performance and forgetting. Finally, we highlight key open challenges, including handling extreme heterogeneity under temporal drift, designing scalable and
privacy-preserving
memory mechanisms, and establishing standardized
benchmarks. This survey aims to serve as a reference and a roadmap for advancing FCL toward robust and deployable real-world systems.
Read the full abstract ▾
All authors Masoume Gholizade, Fabrizio Ruffini, Pietro Ducange, Francesco Marcelloni
Machine Learning
Quantifying Subliminal Behavioral Transfer Ratios in Language Model Distillation
Uwe König, Hamza Kazmi, Ruizhe Li et al. (+1 more)
Distillation of a
language model intended to transfer benign behavior to a student model may also transfer undesirable characteristics, if they are present in the teacher model, a phenomenon known as subliminal learning. While qualitative evidence supports the existence of this effect, its magnitude has not been systematically characterized. This study quantifies subliminal behavioral transfer ratios by steering two teacher models (Llama-2-7B-Chat and Qwen2.5-7B-Instruct) at varying steering strengths and distilling student models using only benign data. Evaluation on 100 JailbreakBench
prompts with GPT-4.1, serving as the evaluator, indicates that transfer is robust but exhibits distinct scaling behaviors. Llama-2 demonstrates a sharp threshold ($\tau = {0.25,0.32} \ \text{beyond} \ \alpha = -0.15$), whereas Qwen2.5 displays continuous and higher levels of transfer ($\tau$ up to $0.61$).
Read the full abstract ▾
All authors Uwe König, Hamza Kazmi, Ruizhe Li, Maheep Chaudhary
Machine Learning
LakeFM: Toward a Foundation Model for Aquatic Ecosystems Using Irregular Multivariate Multi-depth Time Series Data
Abhilash Neog, Sepideh Fatemi, Medha Sawhney et al. (+9 more)
Understanding and
forecasting lake dynamics is critical for monitoring water quality and ecosystem health across lakes and reservoirs. While machine learning methods have been recently applied to ecological time-series data, existing works assume regular sampling in time and depth, and struggle to generalize across lakes with heterogeneous variables, depths, and observation patterns. To address these limitations, we introduce LakeFM, a foundation model for aquatic systems, pre-trained on large-scale ecological
datasets comprising both simulated and observed lakes. Through extensive empirical evaluation, we show that LakeFM learns meaningful representations spanning broader lake-level characteristics, and achieves competitive or often superior-
forecasting performance compared to existing time-series foundation and non-foundation models, while producing physically plausible predictions consistent with real-world lake dynamics.
Read the full abstract ▾
All authors Abhilash Neog, Sepideh Fatemi, Medha Sawhney, Kazi Sajeed Mehrab, Aanish Pradhan, Bennett J. McAfee, Emma Marchisin, Arka Daw, Robert Ladwig, Cayelan C. Carey, Paul Hanson, Anuj Karpatne
Machine Learning
A prior-free blind detection of information leakage from model predictions
Laurence A. Jacobs
Data leakage – contamination of a model with information unavailable at baseline – is the dominant reproducibility failure in machine-learning-based science, yet detection tools require training code, external data, or domain expertise. None operates on the artifact an auditor most often holds: the model's output. We ask what can be decided about leakage from predictions and outcomes alone. We give a decision-theoretic framework in which leakage diagnostics are functionals of the predicted-risk/outcome law, parameterized by a threshold-weighting linked to proper scoring rules and decision-curve analysis. We prove a sharp impossibility: a recalibrated leak matching an honest model's calibration and discrimination is indistinguishable from honest performance by any function of the predictions, so the broad class is detectable only against an externally supplied ceiling on achievable discrimination. We then prove what leakage cannot hide: a near-deterministic subgroup – the signature of a near-label leak – produces a sustained unit-purity head that no legitimate predictor of a non-deterministic outcome can manufacture, yielding a prior-free test. These results organize leakage into a trichotomy – miscalibrated, broad-calibrated, and deterministic – each with a matched detector and failure mode. We validate on UK Biobank using time-windowed comorbidity leakage with known, graded severity, measuring a detection floor of $\Delta\cstar \approx 0.007$ on this endpoint, below which residual leakage is undetectable from output and too small to alter conclusions. The numerical floor is cohort- and endpoint-specific; the structural lesson is general: output-only detection fails where residual leakage is indistinguishable from an honestly stronger predictor. The test returns a verdict on a prediction vector in under a second on commodity hardware.
Read the full abstract ▾
Machine Learning
Seeing Before Colliding: Anticipatory Safe RL with Frozen Vision-Language Models
Samuel Tetteh and Cody Fleming
The cost signal that constrained-RL
algorithms optimize against is almost always reactive: the simulator emits a non-zero cost only after a collision has begun, and the Lagrange multiplier of PPO-Lagrangian grows only after the episode budget has been exceeded. At race speeds, where collisions are instantaneous and irreversible, any safety mechanism that waits for cost to accumulate is structurally too late. We present VLM-Safe-RL, a framework that integrates a frozen
vision-
language model into the CMDP Lagrangian update as an anticipatory cost term. The framework comprises four contributions: (i) Decoupled Dual-Path CLIP, independent reward/cost paths that respect the CMDP's factorization; (ii) VLM-Lagrange, an augmented multiplier update that incorporates a per-step VLM cost as an anticipatory term; (iii) Confidence Gating, a Bayes-optimal weight derived from a logistic noise model on the CLIP margin; and (iv) VLMPPOLag, the composed
algorithm. On Safety-Gymnasium FormulaOne L2, our principal evaluation ($n{=}5$ seeds, $10^{6}$ steps, budget $d_{\text{lim}}{=}25$) VLMPPOLag$+$Conf is the only configuration in our default budget comparison that simultaneously retains substantive return ($J_r{\approx}40$) and holds cost within budget on a majority of seeds; the five constraint-aware baselines (PPOLag, CPO, CPPOPID, CPO-CLG, PPOLag-RND) each fail at least one requirement. The mechanism generalizes to held-out MetaDrive Medium (catastrophe rate $41\%{\to}26\%$, 95% bootstrap CI $[-26,-5]$\,pp) and shows directionally consistent transfer to Bullet Safety-Gym; we report honestly where it does not (MetaDrive Easy/Hard, Qwen2-VL backbone) and trace the Hard failure to a Lagrangian-regulation pathology rather than the VLM signal itself. To our knowledge, this is the first work to use frozen VLM signals as an anticipatory cost term inside the CMDP Lagrangian update.
Read the full abstract ▾
Machine Learning · Cross-listed from math.ST
Geometric bias in eigenspace perturbation under random heterogeneous noise
Fengkai Liu, Ke Wang and Wanjie Wang
Spectral methods rely on the stability of principal eigenspaces under random perturbations. Classically, this is quantified by the Davis-Kahan and Wedin
theorems, which bound the eigenspace error via the operator norm of the noise and the relevant spectral gaps. While sharp for arbitrary deterministic perturbations, these worst-case bounds can be wasteful in the low-rank signal-plus-noise setting, as they fail to capture the interaction between the signal geometry and the noise distribution. We study the spectral perturbation of signal-plus-noise
matrices corrupted by sparse random noise with an arbitrary, inhomogeneous variance profile. Under heterogeneous variances, the empirical eigenvectors suffer a systematic, deterministic
geometric bias invisible to classical bounds. Leveraging the Quadratic Vector Equation (QVE) and fine-grained isotropic local laws, we derive near-optimal, non-asymptotic bounds for the leading eigenspaces in the operator and 2-to-infinity norms. These separate the usual signal-to-noise contribution,
stochastic fluctuations, and structured
geometric bias terms determined by the
alignment between the signal eigenspaces and the row-wise variance profile. We further develop refined rowwise bounds that adapt to the variance-weighted leverage of the signal space, yielding sharper guarantees in delocalized regimes. As applications, we establish strong consistency of adjacency spectral clustering for degree-corrected
stochastic block models with heterogeneous degrees and unbalanced communities, recovering the logarithmic expected-degree scale in the regular balanced case. We also study spectral
embedding for generalized random dot product graphs, showing that the full signal
embedding admits sharp rowwise control, whereas spectral truncation can retain a systematic
geometric bias determined by the omitted signal directions and the variance profile.
Read the full abstract ▾
Machine Learning
PermDoRA – Understanding Adapter Interference in Language Models: Limits of Parameter-Space Geometry
Gowtham Sivaramakrishnan, Sarvesha Kumar Kombaiah Seetha, Kishan Gupta Balaji et al. (+1 more)
Access control in large
language models (LLMs) requires modular mechanisms to enable domain-specific behavior without retraining or cross-domain interference. A common hypothesis is that interference during adapter composition arises from overlap in linear parameter updates, suggesting that enforcing orthogonality or directional independence should improve multi-domain performance. We test this hypothesis using DoRA-RBAC, a hierarchical adapter composition framework based on weight-decomposed low-rank adaptation. We compare conventional Euclidean merging with a geometry-aware Riemannian-inspired merging strategy that approximates the Frechet mean via normalized directional averaging across multiple QA
benchmarks (GPQA, PubMedQA, SimpleQA, WMDP) on LLaMA-3.1-8B and Mistral-7B. Our results show that while single-domain performance matches LoRA, geometry-aware merging provides no consistent advantage over standard averaging in multi-domain settings.Diagnostic analysis further reveals that angular
alignment and orthogonality of adapter updates are weak predictors of composition performance. These findings suggest that adapter interference is not governed primarily by parameter-space geometry, but is instead consistent with interactions in shared nonlinear representations.
Read the full abstract ▾
All authors Gowtham Sivaramakrishnan, Sarvesha Kumar Kombaiah Seetha, Kishan Gupta Balaji, Santhosh Baradwaj Vaduvur Ranganathan
Machine Learning
Loss Landscape Diagnosis for Gradient-Based Gray-Scott System Inversion: Disentangling the Roles of PINN Components
Yan Yang
Gradient-based inversion of reaction-
diffusion systems is typically approached via surrogate models or physics-informed neural networks (PINNs), while the most direct route, backpropagation through the PDE's structure itself, has largely been avoided. We pursue this direct route as a diagnostic probe, backpropagating a steady-state loss through unrolled Gray-Scott simulation to recover its parameters, with no surrogate or neural-network augmentation. Optimization fails to converge, and plotting the landscape directly locates the failure in its geometry – flat plateaus with no gradient signal, bounded by sharp cliffs that align with bifurcation boundaries – a structure that recurs across loss functions and is inherited however the gradients are routed to parameters. Reading this minimal setup as an ablation of PINN, we disentangle each component's role: with the neural network fixed, the residual loss is quadratic in the PDE parameters and yields a smooth landscape, so it alone already avoids the pathology, by implicitly encoding the full PDE dynamics across all initial conditions. The neural network, for its part, cannot repair an ill-posed parameter subspace, and so serves only to complete the observed data – a division of labor not previously made explicit. These findings carry concrete design implications for PINN-type methods and a broader heuristic on when added dimensions actually help.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.CL
Energy-Efficient On-Device RAG on a Mobile NPU: System Design and Benchmark on Snapdragon X Elite
Zhiyuan Cheng and Longying Lai
Retrieval-Augmented Generation (RAG) pipelines are compute-intensive, combining
embedding,
retrieval, reranking, and large
language model (LLM) generation. Running them entirely on-device benefits
privacy, latency, and offline use, but the energy cost of CPU inference is a major barrier. We present what is, to our knowledge, the first end-to-end RAG pipeline that runs all neural stages –
embedding, reranking, and LLM generation – on the Qualcomm Hexagon NPU of the Snapdragon X Elite. Profiling on a Dell XPS 13 laptop, we compare NPU-accelerated RAG against CPU and OpenCL/Adreno GPU baselines on indexing and query workloads. On indexing, the NPU achieves 9.1x higher
embedding throughput and 12.3x less system energy. On a 120-query Wikipedia-passage
benchmark, it delivers 18.1x faster LLM prefilling, 4.0x lower end-to-end query latency, and 4.0x less system energy than the CPU baseline; the same workload on the integrated GPU is 1.7x slower than CPU and uses 6.5x more energy than the NPU. A GPT-4.1 LLM-as-judge evaluation finds NPU answer quality on par with CPU and GPU within evaluator noise (mean 9.32 vs. 8.95 vs. 9.03 on a 1-10 rubric), with 86.7% of queries scoring identically across all three backends. On the Snapdragon X Elite / Hexagon class of laptop SoC, the NPU thus enables practical, energy-efficient on-device RAG without quality regression – a sustainable path toward green edge intelligence that we expect to generalize to comparable mobile NPUs (Apple Neural Engine, Intel NPU, MediaTek APU) as their software stacks mature.
Read the full abstract ▾
Machine Learning
When to Align, When to Predict: A Phase Diagram for Multimodal Learning
Ilay Kamai, Hugues Van Assel, Aviv Regev et al. (+2 more)
Cross-modal
alignment (CA) and cross-modal prediction (CP) are the dominant paradigms for
multimodal representation learning, yet there is no systematic understanding of when each succeeds and when each fails, a gap that leaves practitioners, especially in scientific domains with heterogeneous instruments and multiple levels of measurement, unable to diagnose why standard methods underperform the best single modality. We study both objectives under a spiked signal-plus-noise model with structured cross-modal nuisance correlation, the ingredient that breaks the classical recovery guarantees, and derive separation ratios that expose complementary failure modes:
alignment whitens each modality and fails when nuisance is strongly correlated across views; prediction encodes whatever is cross-predictable through a one-sided whitening, with recovery governed by source-modality quality. The resulting phase diagram partitions
multimodal problems into four regimes — Both, CA only, CP only, and Neither — refined by a recovery count that separates partial recovery from complete failure. We present a data-driven procedure to locate real-world
datasets in this diagram using a small labeled subsample, identifying the preferred objective and prediction direction before any cross-modal training, and identifying when no objective in the CA/CP family can improve on the stronger modality alone. Experiments on synthetic data, stereo-
vision benchmarks, image–caption pairs, and two real scientific domains, astronomy and single-
cell multi-omics, validate the predictions in the nonlinear regime, including both faces of the Neither regime. Code to reproduce the results is available at https://github.com/IlayMalinyak/mm_align_vs_pred.
Read the full abstract ▾
All authors Ilay Kamai, Hugues Van Assel, Aviv Regev, Hagai B. Perets, Randall Balestriero
Machine Learning
A Unifying Lens on Supervised Fine-Tuning Through Target Distribution Design
Tong Xie, Yuanhao Ban, Yunqi Hong et al. (+3 more)
Supervised fine-tuning (SFT) typically maximizes the likelihood of every
token in a demonstrated trajectory. However, an observed
token can be non-unique, noisy, or misaligned with the model prior. Strictly fitting toward this one-hot target may be suboptimal, especially when the pretrained model encodes a rich knowledge prior. In this work, we reinterpret SFT as target distribution design: instead of studying only the loss objective, we analyze the
token-level target that the loss drives the model to match. We introduce the Q-target framework, which decomposes SFT supervision into two explicit choices: (1) how strongly to rely on the observed
token, and (2) how to allocate the remaining probability mass over alternatives. This perspective unifies many existing SFT variants as implicit choices of the target distribution Q. Building on this view, we propose Target-SFT which constructs the training objective directly from the desired target distribution. This method consistently outperforms across the ten
reasoning dataset-model settings evaluated, showing the effectiveness of this target-based approach. Overall, our formulation reveals a more fundamental design principle for SFT training and opens a broader search space for SFT objectives.
Read the full abstract ▾
All authors Tong Xie, Yuanhao Ban, Yunqi Hong, Sohyun An, Yihang Chen, Cho-Jui Hsieh
Machine Learning
EEVEE: Towards Test-time Prompt Learning in the Real World for Self-Improving Agents
Weixian Xu, Shilong Liu and Mengdi Wang
In this paper, we propose EEVEE, the first multi-
dataset test-time
prompt learning framework for LLM
agents, enabling test-time
prompt learning under real-world task streams. Existing methods are largely designed for single-
dataset settings, while real-world applications require models to handle heterogeneous input streams drawn from multiple
datasets, domains, and task distributions, limiting their practical applicability. To mitigate cross-
dataset interference, EEVEE introduces a router that partitions incoming inputs into task clusters and assigns them to suitable
prompt configurations. This design is optimized via a router-
prompt co-
evolution strategy, which employs interleaved router and
prompt learning phases to address their mutual dependency. Experiments across multiple
datasets demonstrate that the framework improves robustness under heterogeneous data streams while maintaining single-
benchmark learning capability and efficiency. Specifically, EEVEE improves average multi-
benchmark scores by 10.38 and 24.32 points over Qwen3-4B-Instruct and DeepSeek-V3.2, surpassing SOTA methods GEPA and ACE by up to 37.2% and 48.2%.
Read the full abstract ▾
Machine Learning · Cross-listed from cs.AI
Semih Kara and Oğuzhan Ersoy
Conditioning a
language model on additional context, such as feedback on a previous attempt, typically improves its response. Self-
distillation trains the model to retain this improvement when the context is not present. The method works by matching the model's output distribution under two settings: a student that sees only the question, and a self-teacher that also sees the context. What the model learns therefore depends on what context the self-teacher receives, yet the design of this context remains largely unexplored. We study context design for self-
distillation by training a solver on feedback from a frozen critic. We compare three conditions: (i) a binary reward (GRPO), (ii) the reference solution, and (iii) a step-by-step critique aligned to the solver's
reasoning trace. Step-aligned critique yields the largest gains, outperforming GRPO by 16.11 points and reference-solution-conditioned self-
distillation by 5.27 points (Avg@12). Per-
token advantage analysis reveals why: step-aligned feedback targets only the
tokens where
reasoning fails, leaving correct behavior intact. Conditioning on the reference solution, by contrast, pressures the model to change its behavior at every
token (even correct steps) because an alternative derivation inevitably differs in phrasing and approach. This suggests that structural
alignment between feedback and the solver's
reasoning is a key driver of self-
distillation effectiveness.
Read the full abstract ▾
Machine Learning
Predicting Future Behaviors in Reasoning Models Enables Better Steering
Evgenii Kortukov, Piotr Komorowski, Florian Klein et al. (+5 more)
Deployed large
reasoning models (LRMs) often behave unexpectedly. Test-time steering controls LRM outputs by intervening on their hidden representations, but it can degrade output quality. We argue that prior steering work implicitly relies on internal features that detect behavior in already generated text. We show that these detection features are poor predictors of future behavioral outcomes, and thus not the natural intervention target. Instead, we train activation probes to predict future behavior likelihoods from intermediate
reasoning steps. These probes predict the most likely behavior with 64%-91% accuracy, revealing a separate type of internal prediction features. Building on these prediction features, we introduce a text-level steering method, Future Probe Controlled Generation. FPCG samples multiple candidate sentences and chooses the best one according to a probe predicting the future behavior likelihood. This enables steering with almost no output quality degradation. FPCG also enables steering in several evaluations where activation steering fails. These results show that distinguishing detection and prediction features enables a more nuanced approach to controlling LRM behaviors.
Read the full abstract ▾
All authors Evgenii Kortukov, Piotr Komorowski, Florian Klein, Paula Engl, Gabriele Sarti, Seong Joon Oh, Sebastian Lapuschkin, Wojciech Samek
Machine Learning
Bellman-sufficient Information Complexity
Yunbei Xu
We introduce Bellman-sufficient information complexity for minimax analysis of sequential decision problems. A Bellman-sufficient state retains enough of the history to close the controlled recursion, while an index $Y=\chi(\Omega)$ specifies the decision-relevant information being charged. The upper bound is a log-penalized Bellman program; the lower bound is a Bellman–Fano comparison along an
algorithm-dependent reference trajectory. If the two values match at a common localization scale and the stated admissibility, calibration, and growth conditions hold, they form an information-risk sandwich. UCB, E2D, and AMS/EBO control or relax the upper Bellman bracket in different ways. For the main application, we give a negative answer to a widely studied form of the GP–UCB minimax-optimality question. For every $0<\alpha<1/4$, we construct one bounded continuous kernel whose minimax regret is $\Theta(T^{1-\alpha})$ along an infinite sequence of horizons, while two globally calibrated GP–UCB rules incur linear regret under one fixed truth. An epochwise finite-marginal action-index AIR Bellman policy, implemented through robust AIR/AMS/EBO control, attains the minimax order. The construction separates realized information from the cost of uniform optimism: many low-value directions inflate the exploration multiplier and change the trajectory. Through the canonical RKHS feature map, it also yields a finite-horizon
polynomial minimax separation for the specified maximal-information-calibrated LinUCB rule. A reproducible experiment illustrates the mechanism.
Read the full abstract ▾
Machine Learning
COGENT: Continuous Graph Emulators with Neural Ordinary Differential Equations for Long-Term Physical Forecasting
Zesheng Liu and Maryam Rahnemoonfar
In this work, we present COGENT, a continuous graph emulator with Neural Ordinary Differential Equations for long-term physical
forecasting on irregular geospatial meshes. COGENT encodes a finite history of system states and associated forcing fields and external forcings with a graph-based history encoder, producing node-wise context vectors that capture both local spatial interactions and temporal
evolution. These context vectors initialize and condition a latent Neural Ordinary Differential Equation whose dynamics are driven by interpolated future forcings and explicit relative rollout time. By modeling the forecast trajectory as a continuous latent dynamical system, COGENT can generate predictions at arbitrary future times rather than being restricted to a fixed temporal discretization. A residual decoder maps the resulting latent trajectories back to future physical states, enabling direct multi-step
forecasting without repeatedly feeding predicted states back into the model. This formulation combines graph-based spatial representation, history-conditioned latent dynamics, and continuous-time rollout in a unified framework for mesh-based physical simulation emulation. In order to stabilize training with long-horizon supervision, we also propose effective rollout-horizon sampling and a progressive rollout-horizon scheduling strategy. We evaluate COGENT on transient ice-sheet simulations generated by the Ice-sheet and Sea-level System Model, demonstrating improved long-range stability over autoregressive graph baselines. These results suggest that continuous graph Neural ODEs provide a promising methodology for scalable physical
forecasting on irregular geospatial meshes, particularly in applications that require stable long-horizon predictions and the ability to query system states at arbitrary times.
Read the full abstract ▾
Machine Learning · Cross-listed from stat.ML
Itô maps for any-step SDEs
Zhengkai Pan, Peter Potaptchik, Wenxi Yao et al. (+2 more)
Recent one-step generative models accelerate sampling by learning deterministic flow maps of the underlying dynamics. These methods rely on learning from ordinary differential equations, leaving open how to define an exact
distillation procedure for
stochastic dynamics. We introduce the Itô map, an any-step
stochastic flow map that takes an intermediate state and Brownian path and predicts future states in a single pass. The Itô map formulation yields novel
estimators for inference-time control by providing cheap, differentiable access to posterior samples. Empirically, Itô maps produce diverse, conditionally valid endpoint samples from fixed intermediate states and support strong steering performance on synthetic and image-generation
benchmarks. These results establish any-step SDE integration as a useful primitive for posterior sampling and
stochastic control.
Read the full abstract ▾
All authors Zhengkai Pan, Peter Potaptchik, Wenxi Yao, Michael S. Albergo, Jakiw Pidstrigach
Machine Learning
Efficiently Learning Drifting Halfspaces with Massart Noise
Mingchen Ma, Guyang Cao, Jelena Diakonikolas et al. (+1 more)
We study the problem of learning a drifting concept in the presence of Massart noise. In this framework, an online learner has access to a history of independent samples whose labels are noisy versions of a target concept that may change from round to round. The goal is to output, in each round, a hypothesis with small prediction error. We study the complexity of this learning problem for the fundamental class of margin-separable linear classifiers (halfspaces). On the positive side, we give a computationally efficient learner achieving error $\eta + \tilde O(\Delta^{1/3}/\gamma)$, where $\eta$ upper bounds the Massart noise rate, $\Delta$ is the drift rate, and $\gamma$ is the margin. Interestingly, in the realizable setting, an adaptation of our techniques yields an efficient learner with an improved error rate over prior work. On the lower-bound side, we provide formal evidence of an information-computation tradeoff, strongly suggesting that our
algorithm's performance is essentially optimal. Specifically, while the information-theoretically optimal error scales with $\Delta^{1/2}$, we prove that $\Delta^{1/3}$-scaling is unavoidable for low-degree
polynomial tests, even in the special case of random classification noise.
Read the full abstract ▾
All authors Mingchen Ma, Guyang Cao, Jelena Diakonikolas, Ilias Diakonikolas