Papers

World–action models (WAMs) promise a unified model that predicts action-conditioned futures, generates feasible actions, and supports planning in imagination. However, existing joint video–action models often use computationally heavy, fixed-horizon backbones ill-suited to streaming inference and stable long-horizon open-loop rollouts. We introduce FLEX-WAM, a Flexible and Efficient Block-Causal World–Action Model for unified simulation and policy inference. FLEX-WAM supports variable-length contexts and non-causal prediction horizons, as well as infinite autoregressive generation frame by frame or block by block. Its block-causal, KV-cacheable architecture combines axial attention and blockwise diffusion forcing to enable efficient real-time rollout and deployment-time latency–throughput tradeoffs without retraining. Joint training can nevertheless produce plausible futures that weakly respond to commanded actions. We address this failure mode by balancing state and action flow-matching gradient contributions across the state–action diffusion-noise grid and regulating world-model sampling using Forward-Dynamics (FD) elasticity, an efficient training-time proxy for action responsiveness. Across simulated and real-world datasets, FLEX-WAM achieves superior multi-step prediction quality and latency while producing stable joint state–action rollouts for thousands of steps. As a joint action proposer and simulator within MCTS, it solves long-horizon PushT and all five OGBench Puzzle-4x4 tasks entirely in imagination. On a bimanual OpenArm-based robot, a single checkpoint jointly serves as a play policy and expected-outcome predictor, enabling real-time identification and collection of model–reality mismatches for future self-improvement.

In this pictorial, we consider how the negotiation of creative boundaries with a co-creative AI system can create moments for personal creative reflection. We ground this in our experiences with Froggi-Draw, a single-initiative co-doodling system that gives users power to decide when and how much an AI "collaborator" (Froggi) contributes to their drawing. From a 1-week pilot study where (n=8) novice and experienced artists doodled daily with the system, we report on ways participants navigated creative risk and uncertainty, and how their usage and perceptions of Froggi shifted over time. In moments of disruption, participants described the system as encroaching on their creative territory. We consider how the design of a supportive, co-creative AI "collaborator" might look like the design of a supportive power dynamic—and finding ways to offer users control to find, reflect upon, and flexibly negotiate the boundaries of that dynamic.

Large language models for code generation often fail on execution, multilingual coverage, and contamination control, especially under frozen backbone constraints. We present CodeForge-MA, a unified framework that improves code synthesis through a multi-agent data forge, execution verified reinforced instruction tuning, and a language conditioned mixture of LoRA adapters. Four specialized agents, Composer, Reviewer, Executor, and Curator, iteratively refine instruction code pairs, validate them with tests, and filter duplicates and benchmark leakage. During training, we combine masked supervised fine tuning with a test driven reinforcement objective to align generations with executable correctness. For the larger model, we use sparse expert routing over low rank adapters to improve cross language transfer while keeping the base model unchanged at inference. Experiments show that joint data, objective, and adapter design yields robust gains across programming languages.

Work by David Reutter and Christopher Schommer-Pries has shown that ordinary smooth TQFTs can distinguish (stable) diffeomorphism classes of closed, connected, even-dimensional manifolds subject to certain finiteness conditions. In particular, simply connected, closed, smooth 6-manifolds with finite $\pi_2$ are diffeomorphic if and only if they cannot be distinguished by ordinary TQFTs. The question of whether or not this result holds for all simply connected closed 6-manifolds was open. We consider TQFTs out of the topological and smooth bordism categories, as well as out of a new bordism category constructed in this paper called the formally smooth (FS) bordism category. We present a pair of simply connected closed 6-manifolds with infinite $\pi_2$ that are not homotopy equivalent, yet are indistinguishable by ordinary TQFTs in all three categories. We extend this counter-example to show that ordinary TQFTs cannot distinguish the homotopy type of simply connected closed manifolds in all dimensions greater than or equal to 6 and non-simply connected closed 5-manifolds. We show that our class of counter-example pairs is spinnable and present a non-spinnable class of counter-examples by introducing a $\mathbb{C}P^2$ connected summand. Lastly, we use more highly connected analogues of our counter-examples to show that, for all $n\geq2$ and $m\geq 2n$, there is a pair of $(2n-1)$-connected closed $(4n+m)$-manifolds that are not homotopy equivalent, yet cannot be distinguished by ordinary TQFTs.

Winkelmann conjectures that every smooth flexible complex affine variety of dimension at least two is a Rosay–Rudin space. We verify this for every positive-dimensional quotient $G/H$ with $G$ a connected complex semisimple algebraic group and $H$ a closed connected reductive subgroup. Weak and strong tameness coincide, every injection between tame discrete sets extends to a holomorphic automorphism, and complements of tame, finite, or empty sets are Oka. Every sufficiently sparse enumerated sequence can be sent to a fixed sequence by a composition of $n$ complete holomorphic flow maps, with $n\le6$; injective self-maps of the fixed sequence admit such a realization with $n\le2$. These flows preserve an invariant algebraic volume form. The proof uses entire interpolation and regular functions invariant under two solvable subgroups to construct the automorphisms. If $G$ is simple of rank at least two and $G/H$ is spherical and admits a closed equivariant embedding into an irreducible module, the normalization bound improves to $n\le4$. We also prove that $\SL_2/N(T)$, where $T$ is a maximal torus, is an RR-space. This quotient is the complement of a smooth conic in $\mathbf P^2$.

Natural exponential families (NEFs) are characterized by their variance functions (VFs), namely by the pair \((V,M)\), where \(V\) expresses the variance as a function of the mean and \(M\) is the mean domain. This characterization naturally raises the inverse question: when is a given function \(V\) the VF of an NEF, and what structural properties of the family can be read directly from \(V\)? Bar-Lev (1987), as part of a more general result for absolutely monotone functions, showed that every nonzero polynomial \[ V(m)=\sum_{j=1}^{r} a_j m^j,\qquad a_j\ge 0, \] is the VF of an infinitely divisible NEF on a positive mean domain. This polynomial class splits exhaustively into the cases \(a_1>0\) and \(a_1=0\). Bar-Lev, Letac and Ridder (2024) proved that, after scaling so that \(a_1=1\), the first case yields a counting NEF supported on \(\mathbb N_0\). We establish the complementary result: when \(a_1=0\), the corresponding NEF is absolutely continuous with respect to Lebesgue measure on the positive half-line. More generally, we prove absolute continuity whenever \[ V(m)=m^2G(m), \] where \(G\) is a nonzero power series with nonnegative coefficients. When \(G\) is a polynomial, the mean domain is the entire positive half-line. We also record several consequences for the cumulant structure and for statistical inference.

Dose optimisation of two-agent combinations in early-phase oncology trials must jointly balance toxicity, efficacy, and pharmacologically meaningful drug exposure, yet existing model-assisted combination designs use only binary toxicity–efficacy outcomes and leave pharmacokinetic (PK) data unused for dosing decisions. We propose PKComb-BOIN12, a utility-based Bayesian optimal interval design embedding a continuous PK exposure metric into two-agent dose optimisation via a PK-guided escalation rule and a PK-sufficiency constraint on the final admissible set. Two contributions distinguish this design from a direct combination extension of existing single-agent PK-informed designs. First, PK-based elimination uses a coordinatewise order-ideal (down-set) removal rule, so that eliminating a combination for insufficient exposure is licensed exactly by the assumed monotonicity of exposure in each agent's dose, rather than an unjustified scalar summary of combined dose intensity. Second, the pre-specified target exposure is treated as transportable-but-verifiable: a hierarchical calibration mechanism updates the assumed monotherapy-derived target using accruing combination-arm PK data, with a formal bound on the resulting selection bias under transportability violations. A time-to-event extension, TITE-PKComb-BOIN12, accommodates late-onset outcomes. A factorial simulation study spanning realistic scenarios, including one calibrated to a published combination trial, decomposes the source of PKComb-BOIN12's gains over its non-PK predecessor and evaluates robustness to non-monotone exposure, PK-variance misspecification, and target-transportability violations. An R Shiny application supports design evaluation and real-time conduct.

Scaling decentralized learning changes not only the number of clients $N$, but also the dynamics of information propagation and consensus. We argue that the effect of increasing $N$ cannot be understood in isolation, because data allocation, topology-dependent mixing, and communication capacity may change simultaneously. We study these coupled effects on CIFAR-10 with $N\in\{10,50,100,200\}$, comparing a degree-two Ring, a Static Random graph, and Local-First Heuristic Evolution (LFHE), a locally adaptive topology process based on friend-of-friend discovery. The Ring provides an analytically transparent failure mode: its Metropolis spectral gap decays as $\Theta(N^{-2})$, implying progressively slower contraction of model disagreement as the population grows. Experiments show that holding the nominal local dataset size fixed substantially reduces the apparent population penalty observed when a fixed total dataset is divided among more clients. The remaining degradation depends strongly on communication structure: Ring enters a high-disagreement regime, whereas Static Random and LFHE remain close to consensus. Increasing LFHE's degree threshold further improves accuracy and consensus, but at a substantially higher model-transmission cost. These results show that decentralized scaling is governed by coupled learning and communication dynamics, rather than by the number of clients alone.

We analyse the data that the Apple App Store sends to and receives from Apple servers. We find that multiple cookies are sent by Apple servers and stored on the handset. Adverts with tracking identifiers are also stored on the handset, and we observe that adverts are selected by Apple servers using GDPR special category personal data such as sexuality, religious beliefs and health. User interactions with the Apple App Store (apps and adverts viewed, buttons clicked, searches made etc) are transmitted to Apple servers alongside identifiers linking this data to the individual user and device. We show that much of this data storage and transmission is not essential for the service requested by the user, namely to view, search and install apps. No consent is sought for any of this data storage and processing, and there is no opt out.

Together with Bourgain's slicing problem and the variance conjecture, the Kannan-Lovász-Simonovits (KLS) conjecture has been a driving force for mathematical developments in high dimensional geometry in recent decades. We present here a proof of this conjecture. The crucial step in the proof is the very recent criterion from Song and Zhang involving high-order derivatives of tilt averages.

Hierarchical reinforcement learning improves long-horizon control by organizing primitive actions around persistent subgoals and assigning credit at multiple temporal scales. Recent hierarchical language agents bring these benefits to interactive tasks by explicitly separating subgoal planning from action execution. We observe, however, that an explicit hierarchy does not by itself determine how stable the resulting temporal abstraction is: the learned boundary policy may replace the subgoal almost every turn, making it effectively transient, or retain a subgoal after it has stopped being appropriate. We call this temporal abstraction instability. We propose Stable Temporal Abstraction via Constrained Optimization (STAC), a constrained boundary-policy optimization method that represents premature replanning and stale persistence as constraint costs. STAC applies the resulting Lagrangian costs only to the sampled boundary decision, leaving the underlying algorithm's rewards, critic targets, subgoal advantages, and primitive-action advantages unchanged. Across two backbones and two benchmarks, STAC improves success over a strong hierarchical baseline by $8.1$ and $7.9$ points on ALFWorld and WebShop with Qwen3-0.6B, and by $23.5$ and $15.8$ points with Llama-3.2-1B-Instruct.

Reasoning errors can propagate into later decisions and memory. This survey synthesizes 312 papers and first-party reports on text-based reasoning hallucinations around three questions: what evidence is observable, what study designs establish, and which corrective actions the evidence supports. UIPCA records unsupported premises (U), invalid inferences (I), dependent reuse (P), visible answer-trace consistency (C), and action-policy failures (A). Across 58 reviewed sources, no comparison establishes that a specified intervention improves reasoning while reducing factual reliability under matched conditions. The synthesis connects diagnosis to verification, repair, selective release, and persistent-state control across memory, tools, and training feedback.

We explicitly compute and parametrize the centralizers of elements, encompassing dege\-ne\-rate and unipotent ones, in the orthogonal, symplectic, and unitary gro\-ups over the real and complex fields, including the indefinite cases. The same approach simultaneously yields the isotropy groups of $H$-skew-symme\-tric, Hamiltonian, and $H$-Hermitian matrices under the adjoint representations of the corresponding classical groups. The resulting groups are conjugate to subgroups of nonsingular block matrices whose blocks exhibit a rectangular block Toeplitz structure.

Aspiring international students across Asian countries increasingly depend on commercial education agents to navigate scholarships, documentation, and visas. Alongside this commercial infrastructure, volunteer-run Facebook groups have emerged. Unpaid admins and moderators, often under their real identities, vet information, screen scams, and guide members through scholarships, visas, and departure logistics. We study one such Bangladeshi group, HigherStudyAbroad: Global Hub of Bangladeshis, founded in 2010. Drawing on semi-structured interviews with 17 volunteer admins and moderators, we found that altruism becomes an organizing logic. It shapes moderators' identities, sustains invisible labor, and grants the group legitimacy against paid agents, even amid new technologies such as AI that could reshape this work. Human judgment remains central to how moderators sustain trust. We discuss implications for HCI's understanding of volunteer labor and migration infrastructure, along with design implications for platforms that depend on unpaid, long-term contribution.

We study $N$ bosons on a three-site ring with nearest-neighbour hopping, density-assisted hopping of the opposite sign, and an on-site interaction, written equivalently as three momentum modes $q=0,\pm1$ with a channel-dependent quartic vertex. As the density-assisted term grows it inverts the band, and exact diagonalisation up to $N=400$ shows a first-order ground-state transition from the $q=0$ condensate to a state with $\avg{n_{+1}}=\avg{n_{-1}}\simeq N/2$; the avoided-crossing gap closes faster than exponentially with $N$. In its symmetric eigenstates this state is fragmented in the Penrose-Onsager sense, and at leading order in $N$ it has the energy of a Gross-Pitaevskii (GP) standing wave, as expected for a few-mode mean-field system. We distinguish two cases by switching off umklapp scattering, which is available because $2\cdot(+1)\equiv-1$ on three sites, and by varying the on-site interaction. Without umklapp, $n_{+1}-n_{-1}$ is conserved, the eigenstates form a rotor tower whose lowest member is a fragmented Fock state, and the transition leads to fragmentation exactly when the $\pm1$ interaction is attractive, as predicted by Nozières' exchange argument with the sign of the interaction reversed; otherwise the system forms a single current-carrying condensate. With umklapp, the standing wave is pinned to three positions, the three momentum sectors become degenerate exponentially in $N$, and coherent admixture of the $q=0$ mode stabilises the standing wave over a wider range; the exact ground state is then a number-squeezed, momentum-projected standing-wave condensate rather than a robustly fragmented state.

We investigate soliton formation and interactions in the non-integrable Gardner-Ostrovsky equation (alias the rotation-modified Gardner equation) with anomalous (positive) dispersion. This equation is relevant to the description of wave dynamics in plasmas and other dispersive media. We show that a special class of solitons, characterized by zero total mass and non-monotonic asymptotics, can emerge from pulse-like initial perturbations. Depending on the initial conditions, these solitons may form regular amplitude-ordered trains, irregular nonstationary ensembles of interacting localized structures, or stationary propagating multi-soliton complexes. We further demonstrate that soliton interactions in the Gardner-Ostrovsky equation are inherently inelastic, leading to the emergence of a dominant "soliton-champion" in closed systems. For example, under periodic boundary conditions, only the largest-amplitude soliton ultimately survives, while all smaller solitons are gradually eliminated through successive interactions. Despite their inelastic interactions, Gardner-Ostrovsky solitons exhibit remarkable robustness.

Across biological and social complex systems, models of epidemics or contagions incorporate complexity through network correlations, adaptive behavior, higher-order interactions, or co-circulating strains. Tracking these mechanisms lead to high-dimensional mathematical systems that resist analytical treatment. We show that a broad class of models admits a systematic reduction via fast-variable elimination. The central insight is that complexity can be captured by fast variables equilibrating much faster than the epidemic, leading to “emergent complex contagion” where the force of infection depends on prevalence. The epidemic transition follows a universal equation and is continuous or discontinuous whether the force of infection decays or intensifies with prevalence. Our results reveal diverse epidemic models share a common skeleton such that mechanisms that share a characteristic form can be macroscopically indistinguishable.

We explain autocorrelation decay power laws omnipresent in texts by self-organized criticality. Specifically, we analyze the recently released KLiCKe keystroke dataset and show that not only the final texts' autocorrelations form a manifold that adheres to a power law with a finite-size scaling, but also the text revisions generate revision-size-dependent restoring dynamics toward that manifold. Thus, human writing appears to dynamically regulate semantic correlations in a text toward a critical state.

We give a simpler proof of the theorem of J.~Tao and R.~Travkin saying that (the neutral block of) the affine Hecke category is the monoidal colimit of finite Hecke categories. The argument is axiomatized to apply to various sheaf-theoretic and group-theoretic contexts. We also prove an extension of the Tao-Travkin equivalence for the whole affine Hecke category as a Morita equivalence.

Spatially balanced sampling designs, such as the local pivotal methods (LPM), improve representation of auxiliary-variable space, but design-based variance estimation is challenging because second-order inclusion probabilities are generally intractable and may be zero or near zero. We propose two approaches for estimating the variance of the Horvitz-Thompson mean and the finite-population variance. The first estimates joint inclusion probabilities by Monte Carlo simulation and uses pairwise independence tests to modify the resulting variance estimators. The second, a model-assisted bootstrap (MABS), uses an interpolating Gaussian process to construct a synthetic finite population and estimates variance by resampling under the original LPM design. In simulations, the modified empirical approach performs well when responses are strongly associated with auxiliary variables, but its performance for the Horvitz-Thompson mean deteriorates as this association weakens. MABS exhibits more stable performance across noise levels and sampling settings, including under LPM2 and non-Gaussian noise. For finite-population variance estimation, however, no method shows a uniform advantage. These results illustrate the complementary strengths of the proposed approaches for inference under spatially balanced sampling.

Visual search is a fundamental cognitive ability. This study investigates whether Multimodal Large Language Models (MLLMs) exhibit human-like difficulty signatures in visual search tasks. We compared search performance of humans (n = 1,250) and MLLMs using identical 2D and 3D stimuli across different set sizes. Both groups showed efficient performance in feature searches, most clearly when the target had a unique color, but performance degradation in conjunction searches as set sizes increased. Additionally, we found strong correlations between human and MLLM error rates ($\rho = 0.82$), which suggests that MLLMs are sensitive to similar objective complexities, such as stimulus heterogeneity. However, differences were found as well: whereas humans invested extra search time to respond accurately on target-absent trials, MLLMs exhibited extreme present/absent response biases in complex searches. We conclude that MLLMs replicate high-level human performance signatures, yet their underlying computations differ significantly.

Research results are prioritised by search engines, large language models and knowledge graphs chiefly through restatement rather than through the originating source. Although the open-access share of annual scholarly output has recently exceeded half, a substantial subscription corpus remains outside what machine readers may legitimately access under prevailing licences. A mechanism is described whereby subscription content is drawn into the open corpus through its citation and restatement in open-access articles—a process termed human mining, by analogy with the text and data mining performed by machines. An optimistic upper bound on what may thereby be inferred by a machine reader confined to the open literature is modelled, and the completeness, latency and licence constraints of that bound are quantified. A majority of subscription articles that have ever been cited is found to be reachable through at least one open citation, and the median interval between a subscription article's publication and its first open citation is shown to have fallen from thirteen years for work of 1990 to one year for work of 2020. The advantage once conferred by direct subscription access to machine readers has thereby been largely eroded, chiefly as a consequence of the growth of open publishing rather than any change in how researchers cite.

Reinforcement learning from human feedback (RLHF) has played a central role in making large language models responsive to human instructions. However, human evaluators often favor flattering or persuasive responses over truthful ones, creating incentives for models to appeal to evaluators at the expense of accuracy. AI safety via debate has been proposed as a way to improve the supervision of language models: in this paradigm, two agents argue opposing positions and challenge each other's claims, potentially exposing falsehoods to the adjudicator. A central premise of AI safety via debate is that truthful arguments are easier to defend than false ones under adversarial scrutiny. In this work, we investigate whether this advantage persists when debaters use rhetorical strategies that exploit biases in human judgment. Inspired by competitive debate, we construct 68 LLM-generated dialogues about detective mysteries with known culprits, spanning four interventions: anchoring, fallacy oversight, pro-jargon, and verbosity. We apply each intervention to either the side advocating for the true culprit or the side advocating for an innocent suspect, allowing us to distinguish influence on adjudication from correctness. In a study with 369 participants, we find that, pooled across bias types, these interventions significantly shift judgments toward the manipulated side. These findings expose a vulnerability in debate-based supervision: human adjudication is sensitive to manipulative rhetorical strategies.

We show how extremal near horizon geometries can be recovered from the non extremal near horizon expansion for smooth stationary black hole families admitting a smooth extremal limit. In Gaussian null coordinates common to the family, we apply the near extremal scaling of the surface gravity together with a radial scaling and a rescaling of Killing time. Although the physical surface gravity vanishes, the scaled horizon can retain a finite surface gravity. The extremal near horizon geometry is obtained when this scaled surface gravity is taken to zero. Under the near extremal scaling and the time rescaling, the quadratic radial term in the temporal metric component can survive, while higher radial orders vanish. Expansion through second order is therefore sufficient and is necessary whenever the quadratic term has a nonzero limiting coefficient in the chosen chart. We derive a common pullback prescription for extracting the near horizon coefficients and recover the known throats of Reissner-Nordström, Kerr, Kerr-Newman, and charged rotating Bañados-Teitelboim-Zanelli (BTZ) black holes. Neutral rotating BTZ provides a case in which the string Carroll expansion already suffices.

The Firefighter Problem models a spreading process (originally a fire, alternatively an infection or rumour, for example) on a graph. A defender saves a single vertex per turn; after each defence, the fire spreads to the unburned and undefended neighbours of all burning vertices. Deciding whether a strategy exists for the defender to protect some targeted number of vertices is computationally hard in graphs in general, but tractable in some restricted cases. Inspired by research into spreadable rabies vaccines for bats, we study a variant of the Firefighter problem in which defence also spreads. Some approximation results are already known for this problem; we provide algorithmic and hardness results, as well as containment results for the infinite $n$-dimensional Cartesian and strong grid graphs.

Running the same language model on different graphics processing unit (GPU) vendors can produce different logits, even when the model weights and inputs are the same. We analyze cross-vendor mismatch in two dense and two mixture-of-experts (MoE) models with five metric families, namely bitwise equality, logit differences, top-K consistency, token agreement, and task accuracy. We trace one source of the mismatch to accumulation order inside vendors' matrix instructions. Upcasting to FP32 reduces the dense model's logit error by 43% at three times the runtime, yet keeping only the MLPs in BF16 retains 94% of this gain at 1.3 times the runtime, so most of the cost of full upcasting buys little. In the MoE models, FP32 and FP16 both lower the probability error but raise the logit error and change expert selection, and FP16 fails in the dense model. An output-head low-rank adapter (LoRA) does not help either, since the final hidden state does not predict the mismatch. The mismatch also carries into training. With every seed fixed, a student distilled from a teacher running on AMD answers 431 MMLU questions differently from one distilled from the same teacher on NVIDIA. Under FP32 upcasting, bitwise equality barely changes while the output distributions move most of the way to the reference, so judging cross-vendor agreement by a single measure misreads both its cost and its gains. Code is available at https://github.com/crova-project/crova.

We study the non-parametric estimation of the roughness of a path from discretely sampled observations. Our approach is based on the concept of normalized power variation of a path. The estimator identifies a variation index by comparing coarse increments with local sums of fine increments. We prove pathwise consistency under Hölder regularity, nondegenerate coarse critical power sums, uniform local bounds on fine critical power sums, and separation of the two sampling scales and obtain a convergence rate for the estimator. Neither convergence of critical variation nor monotonicity of the finite-sample statistic is required. The result applies to fractional Brownian motion under dyadic sampling and to Schauder series with bounded coefficients bounded away from zero. For signed Takagi–Landsberg functions we obtain a uniform asymptotic formula and a quantitative bound for every root of the estimating equation in a fixed parameter interval. For fractional Brownian motion, we obtain asymptotic uniqueness of the estimating root and a polynomial almost-sure error bound. We illustrate the results with numerical experiments for Takagi-Landsberg functions, fractional Brownian motion and the stochastic heat equation.

In Networked Markov Decision Processes, transition dynamics are often unknown and the state–action space grows rapidly with the number of agents. In this setting, Taylor representations naturally approximate $Q$-functions, but a naive order-$n$ expansion over $N$ agents requires $\Theta(N^n)$ coefficients. We justify these expansions under smooth expected future local rewards with controlled derivatives. Under this condition, finite-speed information propagation and discounting imply that local-critic Taylor coefficients decay exponentially with the graph distance to the farthest agent involved. Discarding distant-agent coefficients and marginalizing then yield scalable local Taylor representations with a bound controlled by graph locality. Building on these representations, we propose a scalable model-free actor–critic algorithm, establishing finite-sample critic and near-stationarity guarantees for a linear LSTD critic. We then introduce a more expressive neural TD parameterization. Unlike prior constructive spectral methods, our approach covers settings without access to a known local dynamics map, such as hidden switched linear–quadratic regulation. Across three control benchmarks, our method matches or outperforms spectral baselines while scaling efficiently to large graphs.

We propose a force-based model for social navigation of a tour-guide robot. Social forces due to various factors like obstacles, user position and heading, have been accounted for in the model. We claim that each one of these forces makes the robot more sociable to the user and we design an experimental setup for evaluation. In the experiment, the user follows an autonomous robot to a destination in a known map, while undertaking a few simple sub-tasks in the middle, which serve as distractions. For each participant, we run several rounds of the experiment, each with different forces and a shortest path, A*-search baseline model. Using per-round subjective indicators, we propose to study the effect of our force model on constructs such as: follow-ability, perceived safety, and perceived intelligence.

We develop a perturbative framework for systems coupled quadratically to bosonic environments with quadratic Hamiltonians and linear Lindblad operators. The method applies to both single-mode and multimode environments whose initial states may be non-Gaussian and nonstationary. By combining linear evolution in a doubled operator space with a characteristic-function representation, we express ordered multitime environmental correlation functions in terms of derivatives of the initial characteristic function. These correlation functions enter a recursive construction of the time-convolutionless generator at arbitrary finite order. We illustrate the method with a qubit coupled to a damped mode via a two-photon interaction. Initial Fock and thermal states with identical first and second field moments produce different second-order generators through their fourth moments.