Hybrid language models that interleave attention with recurrent components are increasingly competitive with pure Transformers, yet standard LoRA practice applies adapters uniformly without considering the distinct functional roles of each component type. We systematically study component-type LoRA placement across two hybrid architectures -- Qwen3.5-0.8B (sequential, GatedDeltaNet + softmax attention) and Falcon-H1-0.5B (parallel, Mamba-2 SSM + attention) -- fine-tuned on three domains and evaluated on five benchmarks. We find that the attention pathway -- despite being the minority component -- consistently outperforms full-model adaptation with 5-10x fewer trainable parameters. Crucially, adapting the recurrent backbone is destructive in sequential hybrids (-14.8 pp on GSM8K) but constructive in parallel ones (+8.6 pp). We further document a transfer asymmetry: parallel hybrids exhibit positive cross-task transfer while sequential hybrids suffer catastrophic forgetting. These results establish that hybrid topology fundamentally determines adaptation response, and that component-aware LoRA placement is a necessary design dimension for hybrid architectures.
Speculative decoding accelerates autoregressive inference by drafting candidate tokens with a fast model and verifying them in parallel with the target. Self-speculative methods avoid the need for an external drafter but have been studied exclusively in homogeneous Transformer architectures. We introduce component-aware self-speculative decoding, the first method to exploit the internal architectural heterogeneity of hybrid language models, isolating the SSM/linear-attention subgraph as a zero-cost internal draft. We evaluate this on two architecturally distinct hybrid families: Falcon-H1 (parallel: Mamba-2 + attention per layer) and Qwen3.5 (sequential: interleaved linear and attention layers), with a pure Transformer control (Qwen2.5). Parallel hybrids achieve acceptance rates of alpha = 0.68 at draft length k=2 under greedy decoding, while sequential hybrids yield only alpha = 0.038 -- an 18x gap attributable to how each architecture integrates its components. The property is scale-invariant: Falcon-H1 at 3B reproduces the rates observed at 0.5B. We further show that perplexity degradation from a companion ablation study predicts speculative viability without running speculative
Hybrid language models combining attention with state space models (SSMs) or linear attention offer improved efficiency, but whether both components are genuinely utilized remains unclear. We present a functional component ablation framework applied to two sub-1B hybrid models -- Qwen3.5-0.8B (sequential: Gated DeltaNet + softmax attention) and Falcon-H1-0.5B (parallel: Mamba-2 + attention) -- with a pure Transformer control (Qwen2.5-0.5B). Through group ablations, layer-wise sweeps, positional ablations, matched random controls, and perplexity analysis across five benchmarks, we establish four findings: (1) both component types are essential and neither is bypassed; (2) the alternative component (linear attention or SSM) is the primary language modeling backbone, causing >35,000x perplexity degradation when removed versus ~82x for attention; (3) component importance follows a positional gradient, with early layers being disproportionately critical; and (4) hybrid architectures exhibit 20-119x greater resilience to random layer removal than pure Transformers, revealing built-in functional redundancy between component types. These results provide actionable guidance for hybrid mo
Distributed model predictive control (MPC) is currently being investigated as a solution to the important control challenge presented by networks of hybrid dynamical systems. However, a benchmark problem for distributed hybrid MPC is absent from the literature. We propose distributed control of a platoon of autonomous vehicles as a comparison benchmark problem. The problem provides a complex and adaptable case study, upon which existing and future approaches to distributed MPC for hybrid systems can be evaluated. Two hybrid modeling frameworks are presented for the vehicle dynamics. Five hybrid MPC controllers are then evaluated and extensively assessed on the fleet of vehicles. Finally, we comment on the need for new efficient and high performing distributed MPC schemes for hybrid systems.
As the design of single-component battery electrodes has matured, the battery industry has turned to hybrid electrodes with blends of two or more active materials to enhance battery performance. Leveraging the best properties of each material while mitigating their drawbacks, multi-component hybrid electrodes open a vast new design space that could be most efficiently explored through simulations. In this article, we introduce a mathematical modeling framework and open-source battery simulation software package for Hybrid Multiphase Porous Electrode Theory (Hybrid-MPET), capable of accounting for the parallel reactions, phase transformations and multiscale heterogeneities in hybrid porous electrodes. Hybrid-MPET models can simulate both solid solution and multiphase active materials in hybrid electrodes at intra-particle and inter-particle scales. Its modular design also allows the combination of different active materials at any capacity fraction. To illustrate the novel features of Hybrid-MPET, we present experimentally validated models of silicon-graphite (Si-Gr) anodes used in electric vehicle batteries and carbon monofluoride (CFx) - silver vanadium oxide (SVO) cathodes used i
Video world models have shown immense potential in simulating the physical world, yet existing memory mechanisms primarily treat environments as static canvases. When dynamic subjects hide out of sight and later re-emerge, current methods often struggle, leading to frozen, distorted, or vanishing subjects. To address this, we introduce Hybrid Memory, a novel paradigm requiring models to simultaneously act as precise archivists for static backgrounds and vigilant trackers for dynamic subjects, ensuring motion continuity during out-of-view intervals. To facilitate research in this direction, we construct HM-World, the first large-scale video dataset dedicated to hybrid memory. It features 59K high-fidelity clips with decoupled camera and subject trajectories, encompassing 17 diverse scenes, 49 distinct subjects, and meticulously designed exit-entry events to rigorously evaluate hybrid coherence. Furthermore, we propose HyDRA, a specialized memory architecture that compresses memory into tokens and utilizes a spatiotemporal relevance-driven retrieval mechanism. By selectively attending to relevant motion cues, HyDRA effectively preserves the identity and motion of hidden subjects. Ext
We present a unified description of heavy hybrid hadrons based on a constituent-gluon picture embedded in the Born-Oppenheimer (BO) framework. In this approach, the gluonic excitation is treated as a dynamical quasiparticle with a mass generated by instanton-induced interactions. We propose a simple variational derivation of the BO potentials. The main focus of the paper is the derivation of light-front wave functions for hybrid systems, specifically for the $ccg$ and $qqqg$ cases. We employ both variational methods and numerical solutions of the Schrödinger equation in momentum representation. Using the resulting wave functions, we compute the gluon PDFs for these systems.
This paper proposes a bidirectional rapidly-exploring random trees (RRT) algorithm to solve the motion planning problem for hybrid systems. The proposed algorithm, called HyRRT-Connect, propagates in both forward and backward directions in hybrid time until an overlap between the forward and backward propagation results is detected. Then, HyRRT-Connect constructs a motion plan through the reversal and concatenation of functions defined on hybrid time domains, ensuring the motion plan thoroughly satisfies the given hybrid dynamics. To address the potential discontinuity along the flow caused by tolerating some distance between the forward and backward partial motion plans, we reconstruct the backward partial motion plan by a forward-in-hybrid-time simulation from the final state of the forward partial motion plan. By applying the reversed input of the backward partial motion plan, the reconstruction process effectively eliminates the discontinuity and ensures that as the tolerance distance decreases to zero, the distance between the endpoint of the reconstructed motion plan and the final state set approaches zero. The proposed algorithm is applied to an actuated bouncing ball exampl
In large-scale X-ray computed tomography (CT), matrix-free iterative methods are essential due to the prohibitive cost of explicitly forming the system matrix. In practice, forward projectors and backprojectors are often implemented with different discretizations or accelerations, leading to unmatched projector pairs. This mismatch violates the adjointness assumptions underlying classical least-squares solvers, so the resulting iterations no longer correspond to a true least-squares problem and can exhibit non-symmetric or inconsistent behavior. Prior work has explored Krylov subspace solvers such as AB-GMRES and BA-GMRES to handle unmatched projector pairs, where these methods exhibit semi-convergent regularizing behavior. Under matched conditions, AB-GMRES and BA-GMRES reduce to LSQR and LSMR, respectively. However, in the presence of unmatched projectors, AB- and BA-GMRES have been observed to yield improved reconstruction quality compared to classical least-squares solvers. In this paper, we develop hybrid AB- and BA-GMRES methods that incorporate Tikhonov regularization directly into the Krylov subspace iterations. We also examine the relationship between the proposed methods
This paper proposes a hybrid mono- and bi-static sensing framework, by leveraging the base station (BS) and user equipment (UE) cooperation in integrated sensing and communication (ISAC) systems. This scheme is built on 3GPP-supported sensing modes, and it does not incur any extra spectrum cost or inter-cell coordination. To reveal the fundamental performance limit of the proposed hybrid sensing mode, we derive closed-form Cramér-Rao lower bound (CRLB) for sensing target localization and velocity estimation, as functions of target and UE positions. The results reveal that significant performance gains can be achieved over the purely mono- or bi-static sensing, especially when the BS-target-UE form a favorable geometry, which is close to a right triangle. The analytical results are validated by simulations using effective parameter estimation algorithm and weighted mean square error (MSE) fusion method. Based on the derived sensing bound, we further analyze the sensing coverage by varying the UE positions, which shows that sensing coverage first improves then degrades as the BS-UE separation increases. Furthermore, the sensing accuracy for a potential target with best UE selection i
As satellites have proliferated, interest has increased in autonomous rendezvous, proximity operations, and docking (ARPOD). A fundamental challenge in these tasks is the uncertainties when operating in space, e.g., in measurements of satellites' states, which can make future states difficult to predict. Another challenge is that satellites' onboard processors are typically much slower than their terrestrial counterparts. Therefore, to address these challenges we propose to solve an ARPOD problem with feedback optimization, which computes inputs to a system by measuring its outputs, feeding them into an optimization algorithm in the loop, and computing some number of iterations towards an optimal input. We focus on satellite rendezvous, and satellites' dynamics are modeled using the continuous-time Clohessy-Wiltshire equations, which are marginally stable. We develop an asymptotically stabilizing controller for them, and we use discrete-time gradient descent in the loop to compute inputs to them. Then, we analyze the hybrid feedback optimization system formed by the stabilized Clohessy-Wiltshire equations with gradient descent in the loop. We show that this model is well-posed and
Obtaining accurate probabilistic energy forecasts and making effective decisions amid diverse uncertainties are routine challenges in future energy systems. This paper presents the winning solution of team GEB, which ranked 3rd in trading, 4th in forecasting, and 1st among student teams in the IEEE Hybrid Energy Forecasting and Trading Competition 2024 (HEFTCom2024). The solution provides accurate probabilistic forecasts for a wind-solar hybrid system, and achieves substantial trading revenue in the day-ahead electricity market. Key components include: (1) a stacking-based approach combining sister forecasts from various Numerical Weather Predictions (NWPs) to provide wind power forecasts, (2) an online solar post-processing model to address the distribution shift in the online test set caused by increased solar capacity, (3) a probabilistic aggregation method for accurate quantile forecasts of hybrid generation, and (4) a stochastic trading strategy to maximize expected trading revenue considering uncertainties in electricity prices. This paper also explores the potential of end-to-end learning to further enhance the trading revenue by shifting the distribution of forecast errors.
Medication errors pose a significant threat to patient safety, making pharmacist verification (PV) a critical, yet heavily burdened, final safeguard. The direct application of Large Language Models (LLMs) to this zero-tolerance domain is untenable due to their inherent factual unreliability, lack of traceability, and weakness in complex reasoning. To address these challenges, we introduce PharmGraph-Auditor, a novel system designed for safe and evidence-grounded prescription auditing. The core of our system is a trustworthy Hybrid Pharmaceutical Knowledge Base (HPKB), implemented under the Virtual Knowledge Graph (VKG) paradigm. This architecture strategically unifies a relational component for set constraint satisfaction and a graph component for topological reasoning via a rigorous mapping layer. To construct this HPKB, we propose the Iterative Schema Refinement (ISR) algorithm, a framework that enables the co-evolution of both graph and relational schemas from medical texts. For auditing, we introduce the KB-grounded Chain of Verification (CoV), a new reasoning paradigm that transforms the LLM from an unreliable generator into a transparent reasoning engine. CoV decomposes the a
In pseudo-boolean solving the currently most successful unit propagation strategy is a hybrid mode combining the watched literal scheme with the counting method. This short paper introduces new heuristics for this hybrid decision, which are able to drastically outperform the current method in the RoundingSAT solver.
As information ecosystems grow more heterogeneous, both humans and artificial agents increasingly face a simple yet unresolved question: when seeking knowledge, whom should we ask, and why? Inspired by how people intuitively "read a room", this paper introduces the concept of knowledge affordance (KA) to systematize how agents identify meaningful opportunities for information seeking in hybrid human-AI environments. Rather than introducing a fully formed framework, we propose KAs as declarative, semantically grounded descriptions of what a knowledge source can offer, for which kinds of questions, and with which contextual properties. Additionally, we suggest that KAs are relational, possibly emerging from the interplay between the agent's task, preferences and situational factors. Our contribution is thus a conceptual proposal that connects different research streams, including affordances, semantic web services, knowledge engineering and querying, and mutual intelligibility. We sketch possible research directions to build KA-aware systems that navigate information spaces with greater transparency, adaptability and shared understanding.
Detecting unseen ransomware is a critical cybersecurity challenge where classical machine learning often fails. While Quantum Machine Learning (QML) presents a potential alternative, its application is hindered by the dimensionality gap between classical data and quantum hardware. This paper empirically investigates a hybrid framework using a Variational Quantum Classifier (VQC) interfaced with a high-dimensional dataset via Principal Component Analysis (PCA). Our analysis reveals a dual challenge for practical QML. A significant information bottleneck was evident, as even the best performing 12-qubit VQC fell short of the classical baselines 97.7\% recall. Furthermore, a non-monotonic performance trend, where performance degraded when scaling from 4 to 8 qubits before improving at 12 qubits suggests a severe trainability issue. These findings highlight that unlocking QMLs potential requires co-developing more efficient data compression techniques and robust quantum optimization strategies.
A smart home is realized by setting up various services. Several methods have been proposed to create smart home services, which can be divided into knowledge-based and data-driven approaches. However, knowledge-based approaches usually require manual input from the inhabitant, which can be complicated if the physical phenomena of the concerned environment states are complex, and the inhabitant does not know how to adjust related actuators to achieve the target values of the states monitored by services. Moreover, machine learning-based data-driven approaches that we are interested in are like black boxes and cannot show the inhabitant in which situations certain services proposed certain actuators' states. To solve these problems, we propose a hybrid system called HKD-SHO (Hybrid Knowledge-based and Data-driven services based Smart HOme system), where knowledge-based and machine learning-based data-driven services are profitably integrated. The principal advantage is that it inherits the explicability of knowledge-based services and the dynamism of data-driven services. We compare HKD-SHO with several systems for creating dynamic smart home services, and the results show the bette
Visual Question Answering (VQA) is a challenging task that requires systems to provide accurate answers to questions based on image content. Current VQA models struggle with complex questions due to limitations in capturing and integrating multimodal information effectively. To address these challenges, we propose the Rank VQA model, which leverages a ranking-inspired hybrid training strategy to enhance VQA performance. The Rank VQA model integrates high-quality visual features extracted using the Faster R-CNN model and rich semantic text features obtained from a pre-trained BERT model. These features are fused through a sophisticated multimodal fusion technique employing multi-head self-attention mechanisms. Additionally, a ranking learning module is incorporated to optimize the relative ranking of answers, thus improving answer accuracy. The hybrid training strategy combines classification and ranking losses, enhancing the model's generalization ability and robustness across diverse datasets. Experimental results demonstrate the effectiveness of the Rank VQA model. Our model significantly outperforms existing state-of-the-art models on standard VQA datasets, including VQA v2.0 an
Learning from human feedback has enabled the alignment of language models (LMs) with human preferences. However, collecting human preferences is expensive and time-consuming, with highly variable annotation quality. An appealing alternative is to distill preferences from LMs as a source of synthetic annotations, offering a cost-effective and scalable alternative, albeit susceptible to other biases and errors. In this work, we introduce HyPER, a Hybrid Preference routER that defers an annotation to either humans or LMs, achieving better annotation quality while reducing the cost of human-only annotation. We formulate this as an optimization problem: given a preference dataset and an evaluation metric, we (1) train a performance prediction model (PPM) to predict a reward model's (RM) performance on an arbitrary combination of human and LM annotations and (2) employ a routing strategy that selects a combination that maximizes the predicted performance. We train the PPM on MultiPref, a new preference dataset with 10k instances paired with humans and LM labels. We show that the selected hybrid mixture of synthetic and direct human preferences using HyPER achieves better RM performance c
Hybrid codes are widely used to model ion-scale phenomena in space plasmas. Hybrid codes differ from full particle (PIC) codes in that the electrons are modeled as a fluid that is usually assumed to be massless, while the electric field is not advanced in time, but instead calculated at the new time level from the advanced ion quantities and the magnetic field. In this chapter we concentrate on such hybrid models with massless electrons, beginning with a discussion of the basics of a simple hybrid code algorithm. We then show examples of recent use of hybrid codes for large-scale space plasma simulations of structures formed at planetary bow shock--foreshock systems, magnetic reconnection at the magnetopause, and complex phenomena in the magnetosheath due to the interaction of kinetic processes associated with the bow shock, magnetic reconnection, and turbulence. A discussion then follows of a number of other hybrid codes based on different algorithms that are presently in active use to investigate a variety of plasma processes in space as well as some recent work on the development of new models. We conclude with a few brief comments concerning the future development and use of hy