Papers

We construct and classify $Z_2$ quantum spin liquids (QSLs) of fermionic partons on the dipolar-octupolar (DO) pyrochlore lattice. The DO pseudospin shares the space-group and time-reversal algebra of the effective spin-1/2 Kramers doublet, its projective symmetry group (PSG) yields the same 48 symmetric $Z_2$ classes as the generic fermionic classification. We show, the distinct DO symmetry action, encoded in a set of sublattice-independent operators, selects a physically different set of symmetry-allowed mean-field channels, so that the resulting ansatze, spinon spectra, and observables differ from both the effective spin-1/2 case and the bosonic DO classification. Decomposing the anisotropic XYZ exchange model into all symmetric singlet and triplet hopping and pairing channels and minimizing the mean-field energy self-consistently, we obtain phase diagrams in the ($\tilde{J_x}, \tilde{J_z}$) plane in which only a few $Z_2$ states, several supporting gapless or flat-band spinon spectra, are stabilized. For these states we compute the spinon band structures, the low-temperature specific heat, and the dynamical spin structure factor, and we argue that the power-law specific-heat exponent is a diagnostic of the spinon nodal structure. A near-linear-in-temperature specific heat together with broad continua in the structure factor gives a fermionic-parton account of the gapless, fractionalized thermodynamics reported in DO materials such as $Nd_2ScNbO_7$, complementing the bosonic quantum-spin-ice picture of the Ce-based compounds. The work extends the fermionic-parton classification of DO pyrochlore QSLs beyond the $U(1)$ sector, where $Z_2$ topological order is realized equally naturally by fermionic and bosonic spinons.

Electromagnetic Fields (EMF) mapping is essential for wireless propagation modeling and are fundamental to a wide range of applications, including spectrum awareness, network planning, and integrated sensing and communication (ISAC). However, existing datasets are often limited to spatially separated 2D scenes without real base station (BS) information. To address this gap, we present UrbanEMF, a city-scale EMF mapping dataset constructed from real world urban geometry and physical BS deployments using ray tracing. Unlike conventional single scene-based datasets, UrbanEMF preserves continuous city-scale urban coverage and realistic transmitter location information. It further extends traditional 2D maps by incorporating multiple receiver heights. In addition, both aggregated received signal strength (RSS) maps and path loss maps are provided to characterize complementary aspects of the radio environment. This work provides a realistic and flexible benchmark for developing and evaluating learning-based methods in large-scale urban wireless environments. The code for this work is available at: https://github.com/lemonstudy/UrbanEMF

We construct explicit non-malleable affine extractors for every constant entropy rate, with linear output length and exponentially small error, against any fixed number of affine tamperings without fixed points. For every fixed $0<\eta<1$ and $t$, we also obtain entropy threshold $C_{\eta,t}n/\log n$, output length $\lfloor n^{1-\eta}\rfloor$, and error $2^{-n^{1-\eta}}$ against $t$ tamperings. Our extractors, as well as the directional affine extractors of Li and Zhong (CCC 2024), yield explicit Boolean functions with correlation $2^{-\Omega(n)}$ against weakly read-once linear branching programs of size $2^{\Omega(n)}$. For non-oblivious decision trees, we prove linear depth lower bounds for queries of each fixed degree $r\ge2$. Applying Li's sumset extractor (FOCS 2023) gives depth $\Omega_\delta((n/\ell)\log\ell)$ for growing locality $\ell\le n^{1-\delta}$, where $0<\delta<1$ is fixed. In the same range, directional affine extractors give correlation $2^{-\Omega(n/\sqrt\ell)}$ against local trees of depth $c(n/\ell)\log\ell/\log\log\ell$, for a sufficiently small constant $c>0$. Our extractors derandomize the lossless lifting of Efremenko and Itsykson (STOC 2026). For every fixed $0<\xi<1$, this gives explicit polynomial-size unsatisfiable CNFs on $N$ variables whose $\mathrm{Res}(\oplus)$ refutations of resolution depth at most $N$ require size at least $2^{(1-\xi)N}$. Separately, parity substitutions give polynomial-size CNFs on $N$ variables with polynomial-size ordinary-resolution proofs for which every $\mathrm{Res}(\oplus)$ refutation of size $S$ and depth $d$ satisfies $d\log(2S)=\Omega(N^2)$. This removes the $\log^2 N$ loss in the tradeoff of Itsykson, Podolskii, and Shekhovtsov (CCC 2026).

The Island Formula for Quantum Extremal Surfaces of large AdS black holes, which are made to evaporate by coupling the CFT of the AdS/CFT correspondence to an external system, demonstrates that the space time picture can reproduce the quantum mechanically required Page curve of the CFT/bath system if a region in the interior of the black hole has become entangled with the external radiation by the Page time. In an effective quantum field theory (EFT) description in “nice slice" coordinates, this appears to imply paradoxical entanglement between degrees of freedom that have been causally disconnected since long before the Page time. We will give two arguments that all of the quantum information in the island is in fact concentrated in a tiny space-like portion of the island very close to the black hole horizon at the Page time. The first applies the covariant entropy bound\cite{fsb} (CEB) to the causal diamonds of geodesics originating on the space-like region that becomes the island, but on a time slice just before the Page time. All points on the island except those in a stretched horizon region very close to the horizon have entropy bounds that are tiny fractions of the black hole entropy. The second argument uses standard quantum statistical mechanics combined with the tensor network renormalization group regularization of CFT to argue that the wave function of a typical pure state in the microcanonical evaporating black hole always has most of its probability concentrated in a small range of wavelengths corresponding to a small range of radii in the dual tensor network/AdS slice.

Using a frontier model like Claude Opus as a meta-agent to generate terminal tasks and verifiers for RL training is increasingly common. Yet a runnable Docker image and executable test suite do not guarantee a faithful end-to-end pipeline for terminal agent training. We present a meta-agent pipeline motivated by this gap, diagnosing three classes of failure: benchmark invalidity, harness brittleness, and reward misalignment. Prompt redesign and context extension raise baseline solvability 5.6 times, but a 9B model saturates at 81.3% mean pass@2 within 20 steps on Claude Opus-generated tasks. Adding hard tasks reduces mean pass@2 to 20.6% without changing the training configuration, a strong evidence that the solvability band is model-specific. These findings demonstrate that meta-agent reliability requires solvability-band calibration, verifier audits, and infrastructure error accounting as first-class evaluation criteria, not post-hoc diagnost.

We study long context language models. Instead of training long context natively, or designing a long context harness, we train a model over the simplest possible harness: a tool to call itself with any specified prompt and a tool to read tokens in a range from the input context. We finetune Qwen3.6-35B-A3B on a diverse synthetic dataset using this harness. With only 8,000 tokens of context, our small model is as strong as GPT-5.4 with 1M tokens of context on the OOLONG-synth benchmark when document length exceeds 40K tokens.

Energy saving has become a critical challenge in 5G and beyond networks. The rapid growth of connected devices has increased the overall network energy demand, driving operational expenditure to unsustainable heights. The Base Station (BS) accounts for the largest share of energy usage, typically consuming around 60-70% of the Radio Access Network (RAN)'s total energy. Therefore, to address this issue, this article optimizes the BS's energy usage while accounting for the dynamic behavior of User Equipment (UE). Deep Reinforcement Learning (DRL) is a natural candidate for determining effective energy saving policies, such as automatically switching BSs on or off when user density is low or adjusting transmission power to balance energy efficiency and Quality of Service (QoS). However, its heavy training burden and the exponential growth of state and action spaces in dense 5G environments make exploration increasingly difficult. To overcome these limitations, we introduce a novel Quantum Reinforcement Learning (QRL) algorithm that leverages quantum principles, including superposition and entanglement, through parameterized quantum circuits, enabling significantly faster convergence than DRL, which relies on conventional deep neural networks. Extensive simulations demonstrate that the proposed QRL can substantially reduce energy consumption while maintaining QoS, even when UEs are highly dynamic and frequently switch their association with BS antennas. Additionally, QRL consistently outperforms DRL and Q-Learning in both convergence speed and learning complexity.

Flutter onset is determined by the sign of the aerodynamic work on the structure, but cycle-integrated work does not reveal where this energy transfer originates. We perform direct numerical simulations of incompressible flow over an aspect-ratio-two NACA 0015 wing pitching at $Re=1000$ and use lobe-resolved force-moment partitioning to identify the vortical contributions to aeroelastic power. The net energy transfer changes sign near $U^*=4.25$. Across this transition, the upper leading-edge contribution changes from damping at $U^*=2$ to excitation near onset and becomes the largest positive contribution at $U^*=6$. Near the stability boundary, positive leading-edge work is nearly balanced by negative trailing-edge work. This change is accompanied by a reorganization of the leading-edge pressure footprint and a reduction in the phase difference between aerodynamic moment and pitch rate, linking flutter onset to the timing of the leading-edge vortex system. A single slowly growing-amplitude simulation recovers the finite-amplitude energy boundary, while a free-pitching calculation confirms the predicted instability. As the motion approaches its limit cycle, the energy pathways redistribute, and a response-frequency-matched forced calculation reproduces the large-amplitude energy budget. These results distinguish the mechanism that triggers flutter from that which saturates the response and identify the leading edge as a natural target for flutter suppression.

Clock meshes are used in high-performance VLSI designs to minimize skew and tolerate on-chip variation, but they spend scarce routing resources on premium metal layers. Backside power delivery creates a new option: it adds thick, low-resistance metal layers on the back of the wafer, and the power grid does not consume all of them. Flip-flops remain on the frontside; a backside mesh therefore cannot drive them directly, and every connection passes through a through-silicon via. We present the first design-space exploration of backside clock meshes, implemented in OpenROAD on GT2N, a 2 nm nanosheet technology. Four benchmarks (1,938 to 15,311 flip-flops) are explored with multi-objective Bayesian optimization, and every design point is verified by transistor-level SPICE simulation, since the cyclic mesh cannot be evaluated by static timing analysis. Across all four designs, the backside mesh consistently outperforms an identical frontside mesh, with on average 45% lower skew, 25% lower sink slew, 4.5% lower power, and 28% less frontside clock wiring, and its skew spread under 10,000-sample Monte Carlo is a third of the frontside mesh's.

We reveal that a chiral cavity distinguishes the resonant and antiresonant interactions between a two-dimensional electron gas and the cavity vacuum. For co-rotating helicities, the modes anticross, exhibiting vacuum Rabi splitting, and retain a bare vacuum ground state. For counter-rotating helicities, the modes cross, acquiring vacuum-mediated energy shifts and a two-mode squeezed ground state. We trace the origin of these observations to the SU(2) and SU(1,1) symmetries of the respective Hamiltonians and present their spectroscopic signatures in chiral photonic crystal cavities. We further predict that cavity chirality produces a handedness-dependent correction to Hall resistivity at finite polariton broadening, while such a correction is not present in an achiral cavity. For the longitudinal resistivity, we find an asymmetric response of the Shubnikov–de Haas oscillations for the different helicities. Finally, we present an experimentally accessible route for detecting the squeezed ground state through current fluctuations. Parameters based on a recently demonstrated terahertz chiral cavity place these effects within experimental reach.

Multimodal agents are increasingly used for data visualization tasks but remain limited in autonomous review. Unlike humans, they may fail to recognize when a visualization is incorrect, determine what to change, repair it without disrupting correct content, and verify whether the intervention succeeded. Existing benchmarks largely evaluate predefined individual capabilities such as chart generation, instruction-guided editing, or defect detection, and therefore do not capture this gap in autonomous review. We introduce VisAudit, a benchmark for evaluating visualization diagnosis, repair, and verification. Given a rendered chart and configurable auxiliary evidence, including its source data table, intended text summary, and visualization code, an agent iteratively diagnoses potential defects, modifies and executes visualization code, inspects execution and visual feedback, and determines when no further intervention is needed. VisAudit defines three tracks spanning diagnosed repair, autonomous repair, and open-world verification, and contains 1,900 flawed instances across 21 chart types and 10 flaw categories, together with 300 initially correct charts. We construct the benchmark through controlled perturbations of validated source visualizations, with systematic verification and human-aligned quality control to ensure that injected defects are well-defined and recoverable from the available evidence. Experiments with leading multimodal models reveal a substantial gap from reliable autonomous review: the strongest evaluated model fully recovers only $47.4\%$ of flawed charts in the autonomous-repair setting.

A robot demonstration records two things in the same frames: what happened to the objects, and how one particular arm made it happen. We condition on the first. A demonstration is compiled into an effect program: the 3D keypoint trajectories of the objects that moved, two points marking where each was held, and the configuration the scene ends in, with the demonstrator removed. PEWAM, a 71.5M-parameter world-action model, generates effect, robot execution, action and terminal state as four streams with independent flow-matching times, so clamping a program and sampling the execution turns inference into programming, re-solved closed loop from the live scene. On held-out LIBERO-Goal tasks, one demonstration's program completes 40 of 90 episodes, where the same backbone given a goal image or language, and published demonstration-conditioned methods, complete at most 19; on three of Meta-World's held-out classes it exceeds the best published results, though not on the five-class mean. Because a program is a set of coordinates, a person can edit it: the placement follows a shifted terminal state and the grasp turns with rotated contact points. The same program runs on four robot arms without retraining, and after a push, re-solving completes 23 of 60 episodes where replaying the demonstration completes 6. On a Franka arm, fine-tuned on real demonstrations of other tasks, programs compiled from single human videos complete 36 of 40 trials, against 22 for the same backbone conditioned on the video's last frame as a goal image.

Electric utilities and grid operators increasingly rely on machine-learning models to forecast next-day demand, and those models learn from meter data that is routinely defective: readings go missing, sensors freeze, buildings read zero for hours, and units change by a factor of 100. This report presents a data-onboarding pipeline that detects and repairs such defects before a model is trained, using only information available at forecast time, and a controlled experiment that measures whether the pipeline protects a 24-hour-ahead forecast. On hourly electricity data for twelve U.S. buildings from the public Building Data Genome 2 dataset (210,528 rows, 2016-2017), seeded, hash-logged defects touching 0.10% of the training period raised the error of a gradient-boosting forecaster by 86%; after detection and past-only repair the error returned to the clean-data level (mean absolute scaled error 0.760 clean, 1.415 corrupted, 0.729 repaired) while 93% of training targets were retained. At a defect prevalence calibrated to published field studies (1.6% of training rows) the unprotected forecaster's error reached 4.4 times that of a seasonal-naive rule, and the repaired forecaster again matched the clean baseline. The same pattern held for ridge regression and a random forest and across horizons of 1 to 24 hours. A first version of the pipeline over-cleaned natural data and made forecasts 25% worse; that result is retained, its cause is traced in the published artifacts, and the per-building calibration that corrects it is documented as a dated amendment. Every number is reproducible from pinned public inputs with SHA-256 verification, 84 automated tests and continuous integration.

Multi-agent systems (MAS) built from large language models coordinate specialized agents to tackle complex tasks, but effective workflows are difficult to design in advance. Test-time evolution refines workflows using execution feedback, yet broad revisions can disturb useful components, while re-executing unchanged requests can incur redundant computation. Inspired by the interplay of inheritance and selection in biological evolution, we introduce Inherit-MAS, which makes inheritance explicit at the workflow and execution levels. A meta-model first synthesizes a workflow of worker agents with declared roles, communication inputs, and tool permissions, and a separately prompted judge scores each executed candidate and diagnoses its deficiencies. In ordinary refinement rounds, workflow inheritance starts from the latest completed candidate, may discard removable nodes judged unhelpful, and applies a validated edit to address the diagnosed deficiency. When the new candidate executes, execution inheritance inherits eligible stored results only if the complete resolved request and execution context match, avoiding redundant model and tool calls. With GPT-4o-mini workers, Inherit-MAS achieves 55.4% completion on WorkBench and 49.7% joint F1 on HotpotQA FullWiki, outperforming EvoAgent, EvoMAS, and TacoMAS. With Qwen3-32B workers, it also exceeds these evolving-MAS baselines on both benchmarks. Compared with rerunning the same controller with execution inheritance disabled, execution inheritance reduces worker-token usage by 29.1% on WorkBench and 34.6% on HotpotQA, and total token usage by 5.3% and 18.1%.

Streaming GPU solvers for entropic optimal transport (EOT), such as FlashSinkhorn, avoid storing the dense kernel but still evaluate all $n\times m$ point pairs in every Sinkhorn iteration. We present FlashSinkhorn~2 (FS2), a solver for squared-Euclidean cost on low-dimensional point clouds that solves large discrete EOT problems to a prescribed marginal residual on a single GPU by coupling two stages. A coarse stage solves on cell centroids, lifts the potentials to every point and, when a sampled marginal check rejects the lift, continues on the centroids, replacing most point-level updates. A block-sparse fine stage then removes the centroid error that coarse updates cannot. Its Morton-ordered blocks support screening and fused tensor-core execution, and a threshold set by the block masses bounds each omitted tile's contribution to every row and column. On synthetic benchmarks, FS2 reaches the target residual on all 32 problems and GeomLoss multiscale on 10. On one A100, FS2 solves discrete EOT between two $1.34\times10^8$-particle measures from a cosmological $N$-body simulation, at an entropic blur equal to the mean interparticle distance, to an all-particle marginal residual below 0.01 in under 2.5 hours. To our knowledge, it is the largest discrete EOT problem solved to this accuracy within hours. For reproducibility, we release an open-source implementation at https://github.com/ot-triton-lab/flash-sinkhorn

We apply the stochastic all-mode technique to general coarse-graining schemes of the tensor renormalization group (TRG) approach. The technique possesses a notable characteristic of being entirely free from systematic errors which can potentially arise in the coarse-graining step, while it has statistical errors caused by random noise. The key element in our generalization is employing a method called squeezer, and the all-mode technique can be used when it is constructed. Since many TRG methods can be described in terms of the squeezers, our approach can be applicable in all such cases. In this work, we demonstrate the new method with higher-order TRG in the two- and three-dimensional Ising model.

We undertake a systematic study of multiplicative monoid morphisms between prime-power modular monoids $(\mathbb{Z}/p^e\mathbb{Z},\cdot)$ and $(\mathbb{Z}/q^f\mathbb{Z},\cdot)$, for arbitrary primes $p,q$ and integers $e,f\geq 1$. Our approach proceeds in three stages. We first determine the hom-set $\mathbf{Hom}(U_{p^e},U_{q^f})$ of group morphisms between the corresponding unit groups and exhibit explicit isomorphisms describing its group structure. We then identify the subset $\mathcal{H}_{p^e,q^f,v}$ of unit group morphisms that extend to full monoid morphisms on $(\mathbb{Z}/p^e\mathbb{Z},\cdot)$, characterizing it via explicit divisibility conditions and describing its algebraic structure. Finally, we give a complete description of $\mathbf{Hom}_{\mathbf{Mon}}((\mathbb{Z}/p^e\mathbb{Z},\cdot), (\mathbb{Z}/q^f\mathbb{Z},\cdot))$ and establish an explicit monoid isomorphism between this hom-set and a naturally defined monoid built from unit groups of lower-level modular rings.

We propose Q(q,d), a family of q-ary non-CSS qudit stabilizer codes with prime power q and dimension d >= 2, constructed from Singer difference sets and non-degenerate quadrics in PG(d,q). The parity check H = (A | M_Q A) couples the Singer-circulant incidence matrix A with the quadric circulant M_Q, and a Cancellation Lemma uses the identity (k - lambda) = q^{d-1} = 0 mod p to make H symplectic-isotropic over F_q. Five structural theorems carry the novelty: a cross-correlation modular-constancy identity nu = 1 mod q at q odd, d >= 3, with a two-value pattern that proves a 2-torsion distance ceiling; a CRT zero-count identity; an explicit Clifford bridge identifying Q(q,2) as a graph-state encoding of the trivial Singer CSS code, with the 2021 low-density-spreading (LDS) sign-flip pattern realised by a symmetric circulant; a paired column-rigidity equality d_min = R_P + 1; and a lower bound d_min(Q(7,2)) >= 8 from an exhaustive Singer-Frobenius search. A Hamada-Smith specialisation yields the closed-form logical dimension k(p,d) = theta_d(p) - C(p+d-1, d) - 1, and the flagship Q(p,2) = [[p^2 + p + 1, p(p+1)/2, d_min]]_p attains asymptotic rate one-half. We conjecture d_min(Q(p,2)) = p + 1 for odd primes, proven at p = 5 and lower-bounded to >= 8 at p = 7. Instances include Q(3,2) = [[13,6,3]]_3, Q(5,2) = [[31,15,6]]_5, Q(7,2) = [[57,28, d>=8]]_7, and Q(3,4) = [[121,105,3]]_3. Monte-Carlo over 1.5 x 10^6 trials shows Q(5,2) achieves a 70-fold per-logical-qudit improvement over Steane [[7,1,3]]_2 at depolarising rate 10^-3 with zero silent logical errors. A textbook bounded-distance decoder is used; Berlekamp-Massey-style cyclic decoding is left as the natural sequel.

When a large language model solves a mathematical problem, its reasoning is largely hierarchical, and the solution often branches at a few tokens where the next-token entropy is high. Such tree-like structure embeds in hyperbolic space with far lower distortion than in Euclidean space. Activation steering, however, usually edits the hidden states of a pretrained model by adding one fixed Euclidean vector at every token, even though most tokens of a solution are already determined by the context. We propose Hyperbolic Entropy Steering (HEST), which embeds the hidden states in the Poincaré ball with a lightweight probe whose only label is the model's own next-token entropy. Where this entropy exceeds a threshold, HEST moves the embedded state along the geodesic of steepest descent of a readout of the probe and maps the change back to the hidden state. For the Busemann readout of a learned ideal point, we prove that a step of fixed length lowers it by the same amount at every state. On three instruction-tuned models from the Qwen2.5-Math and Llama-3.1 families, HEST with the Busemann readout improves greedy accuracy on MATH-500 and GSM8K in five of six settings, by up to 1.8 points, whereas a contrastive steering vector added at every token lowers accuracy. With a Euclidean probe trained in the same way, this gain disappears on Qwen2.5-Math-1.5B-Instruct. The gains are largest on problems where the model hesitates often, and accuracy on the remaining problems is almost unchanged.

We construct near-spanning balanced antidirected trees that fail to embed in highly symmetric tournaments despite the conjectured semidegree condition being satisfied. More precisely, for every sufficiently large odd integer $n$, we exhibit a regular vertex-transitive tournament on $n$ vertices and a balanced antidirected caterpillar on $n-1$ vertices with maximum degree at most $(1+o(1))n/\log_2 n$ that is not contained in the tournament. The host satisfies the strict $k/2$ threshold in both the semidegree and pseudo-semidegree settings, where $k$ is the number of arcs of the target tree. Consequently, this gives counterexamples to the uniform sublinear-degree interpretations of Conjectures~6.8 and~7.6 in Stein's survey (2024). The order $n-1$ of the target is best possible under the strict semidegree hypothesis. We also give a six-vertex balanced antidirected double-star missing from the seven-vertex Paley tournament, and show that seven is the smallest host order for a pseudo-semidegree counterexample with this fixed double-star.

Here, we explore magnetic field amplification by the small-scale dynamo (SSD) during the formation of the first stars and galaxies. Gravitational collapse continuously modifies turbulent velocities and length scales, making the SSD fundamentally different from a dynamo operating in stationary turbulence. In arXiv:2503.19131, we developed a supercomoving formulation of turbulent dynamos and showed that the increasing dynamo growth rate during collapse leads to super-exponential (SE) magnetic field amplification in the kinematic stage. Here, we show that collapse also accelerates the slower nonlinear stage, allowing magnetic energy to reach progressively larger scales. We apply this framework to first-star and first-galaxy formation and find that the impact of collapse depends strongly on the dynamo efficiency. If the dynamo is sufficiently fast, the field reaches saturation before substantial collapse occurs, leaving little room for further acceleration. At the opposite extreme, if the dynamo is too slow, dynamo amplification remains weak and flux freezing dominates. Between these limits, collapse can substantially accelerate the dynamo, with speed-ups of >10 in the parameter space explored, enabling dynamically significant fields to develop within the available collapse time. Nonlinear scale-by-scale saturation further produces fields correlated on a non-negligible fraction of the system. We provide estimates and scaling relations for the final field strength and correlation scale and their dependence on halo, turbulence, dynamo, and collapse parameters. The earlier emergence of dynamically important fields can affect fragmentation, angular-momentum transport, and the subsequent evolution of the first stars and galaxies.

Existing 3D large language models (LLMs) compromise on two fronts: they compress shapes into latent codebook indices or coordinate text, which removes spatial structure from what the model observes, and they acquire the 3D modality by fine-tuning the backbone, which overwrites its general language ability. We present OctLLM, which addresses both limitations. Geometry enters as an explicit 3D sequence of octree occupancy tokens. However, full octree sequences grow rapidly with depth; OctLLM therefore randomly empties penultimate-level nodes and omits descendants while preserving shape, yielding a shorter coordinate- and depth-anchored Sparse Octree (S-Octree) for position-aware mask-modeling generation and 3D understanding. On the other front, existing methods introduce a new modality with full fine-tuning or LoRA, but full fine-tuning is costly, LoRA limits 3D capacity, and both modify the language pathway. OctLLM instead adds 3D capacity in parameters separate from the pretrained ones: mesh tokens are routed through independent trainable branches in a subset of blocks while text and image tokens retain the frozen vision-language pathway, and the two streams interact through shared self-attention. It trains far fewer parameters than full fine-tuning, yet sets a new state of the art among unified multimodal LLMs, lowering image-to-3D FID by $17.4\%$ and raising render-grounded captioning by $28.7$ points over ShapeLLM-Omni, while matching the backbone on general language benchmarks.

In every fast nearest-neighbor index, recall is measured, never predicted: each operating point is tuned by serving it against ground truth. SOLO is an index whose recall is computed from the index itself, before any query is served. SOLO is an inverted file whose vocabulary is a random sample of the database. Each object's $k_b$ nearest sample points are stored once, ranked; at serve time an object is posted under the first $b \le k_b$ of them, and a query scans the lists of its $k_s$ nearest sample points exhaustively, with the true distance. Recall depends on the product $b \cdot k_s$ (an equal-work law), so the search-side $k_s$ compensates for a small $b$ with no rebuild. There is no beam, vote or pruning bound, so a true neighbor is missed only if it shares no sample point with the query – a membership event decided by stored integers, not by a search. Recall is therefore a count: one ground-truth pass over a sample of the operator's queries certifies every $(b, k_s)$ at once, with nothing served. The certificate matches served recall to four decimals from $10^6$ to $10^9$ objects, including 768-dimensional text embeddings under inner product with shifted queries. No graph index has an analogous object. The rest is the same rule applied recursively: a list that outgrows a bound is sampled and split like the database, and so is the vocabulary itself, which is what drives resident memory down. Deep-100M is served at recall 0.9977 from 1 GB of enforced resident memory, Deep-1B at 0.9925 from 96 MB. Inserts are one search, deletes are exact, and throughput reaches $1.8\times$ a tuned HNSW at $10^8$. Every number is reported against HNSW, DiskANN, GRAFT, NAPP, misi, SPANN, ScaNN and RaBitQ on the same hardware and ground truth.

We present a comprehensive overview of the challenges and opportunities in social bot detection in the context of the rise of sophisticated AI-based chatbots. By examining the state of the art in social bot detection techniques and the more salient real-world application to date, we identify gaps and emerging trends in the field, with a focus on addressing the unique challenges posed by AI-generated conversations and behaviors. We suggest potentially promising opportunities and research directions in social bot detection, including (i) the use of generative agents for synthetic data generation, testing and evaluation; (ii) the need for multimodal and cross-platform detection based on network and behavioral signatures of coordination and influence; (iii) the opportunity to extend bot detection to non-English and low-resource language settings; and, (iv) the room for development of collaborative, federated learning detection models that can help facilitate cooperation between different organizations and platforms while preserving user privacy.

In recent years, several variants of classical hash table schemes have been developed by engineers in order to take advantage of the processor's internal parallelism using SIMD instructions, which make it possible to operate on multiple bytes simultaneously. At a small additional memory cost, this enables a significant speedup, making it a data structure that is increasingly popular in practice, when very high performance is required. In this article, we provide a detailed theoretical analysis of the dynamics of such hash tables. From a methodological standpoint, we use and adapt a technique developed by Wormald in the 1990s to study dynamic graphs. This approach, which can be adapted to many variants, enables us to accurately estimate the quantities of interest by capturing the dynamics of the data structure through systems of differential equations. Although complex, we provide an explicit description of the solutions of these systems, which can furthermore be efficiently approximated numerically. Our main results are stated with high probability, which is significantly more precise than average-case analyses, and they match experimental results remarkably well, even for hash tables of moderate size.

AI-assisted cyber situational awareness triggers machine-generated reasoning traces (step-by-step justifications for anomaly classifications) that a human operator is expected to review. Because cyber signals and their traces arrive faster than any operator can process, human review is the limiting constraint on oversight. The standard approach is to identify the riskiest cyber events for review using model-side signals such as confidence or uncertainty. That framing ignores the operator's cognitive capacity which varies sharply with the operational environment. We propose an alternative where the system's environmental and connectivity telemetry serves as an available, non-invasive proxy for operator load. That same telemetry determines whether the human-AI partnership can reach the broader collective for support. In a maritime platform, environmental and connectivity attributes including depth, number of active communications paths, density of the tracked contact picture, and operational tempo all carry this signal. Need for operator oversight becomes a decision that materializes as a combination of both risk and environment-derived operator capacity. We present a reference architecture for a connectedness-aware oversight engine, demonstrating everyday use cases alongside its intended incorporation into the submarine cyber-defense toolkit. Two themes emerge: 1) the operator's environmental state is itself a connectedness measurement, and 2) connectedness drives the cognitive load and defines a collective boundary in human-AI cyber operations.

Looped Transformers apply the same layers several times per token, adding compute to improve quality without more parameters. Each recursion, however, writes its own key-value cache, so memory still grows with compute. Inference-time techniques can shrink this cache at a cost in quality. We pretrain looped language models to share memory: only the first recursion writes a cache, and later recursions read it while keeping a short window of their own. Surprisingly, we find that sharing memory does not cost quality and instead improves it. At 150M-1B parameters, our Looped Prediction Transformer (LPT) and its hybrid variant set a new quality-memory frontier for looped models: with five recursions, the hybrid lowers validation perplexity on FineWeb-Edu by 1.12-1.82 relative to a same-size standard Transformer while using 76-79% less context memory. Through an extensive analysis, we investigate why memory sharing helps. Shared and local memory develop different representations, and later recursions attend mostly to the shared memory, which also acts as a gradient highway to the first recursion.

Transfer functions (TFs) control color and visibility in medical volume rendering, but image-trained Gaussian proxies typically bake one transfer function into their appearance. We present FactorSplat, a per-scene N-dimensional Gaussian splatting (N-DGS) proxy that accepts region-specific intensity-to-RGBA curves at inference. A local lookup applies the authored color and opacity change, while a shared functional encoder and low-rank per-Gaussian factors learn the residual appearance response. Geometry and directional appearance remain shared across presets, with visibility control and TF-aware pruning preserving the ability to hide and reveal structures. On seven CT and MR scans, FactorSplat improves mean PSNR and changed-region error over region-aware VEG across validation, interpolation, unseen composition, and out-of-distribution (OOD) edits. Across these four splits, seven-scan mean PSNR gains over VEG range from 1.10 to 1.52 dB. One checkpoint per scan supports unseen edits without retraining. At $1600^2$, the cached fast renderer averages 524 FPS with 1.17 ms TF switches. Project page: https://gaozhongpai.github.io/FactorSplat/.

On-policy distillation (OPD) trains a student on the responses it generates. Existing LLM multi-teacher OPD transfers what specialists predict through their output distributions. We introduce Latent-MOPD, to our knowledge the first representation-level multi-teacher OPD method for LLMs. It integrates existing specialists through both their predictions and the hidden states used to compute them, without additional teacher training. To coordinate representation supervision from multiple specialists, we select late-layer targets according to the teacher-student relationship, bridge unequal hidden widths with a shared projection, and group updates by domain. Each teacher's supervision gradually shifts from hidden states to token predictions, with both channels using the same routed specialist. In our main same-family setting, Latent-MOPD outperforms the token-only, representation-only and uniform-averaging baselines on all nine benchmarks across math, code and logic. With the same parameter count as each teacher, the student also surpasses the per-benchmark best teacher on a majority of these benchmarks. With larger, separately developed cross-family teachers, Latent-MOPD outperforms both single-channel baselines on all benchmarks. A same-family all-layer representation-only control remains stable with domain-pure updates but collapses when teacher domains are interleaved within an update. Our results show that a single student can integrate capabilities from several specialists through both their output distributions and internal representations.

Ambient-temperature coexistence of ferroelectricity, photoconductivity, and ionic transport in ultra-thin van der Waals (vdW) compound CuCrP2S6 (CCPS) offers significant potential for multifunctional electronic, photonic, and neuromorphic applications. However, ferroelectricity and ionic transport can become antagonistic when sharing a common atomic origin, like the copper ions in CCPS. We address this fundamental dichotomy by systematically investigating the thickness-dependent evolution of ferroelectricity, ionic migration, and memristive behavior in CCPS diodes via electrostatic and photo-assisted interfacial band tuning. Using piezoresponse, Kelvin probe, and conductive atomic force microscopy, alongside photoexcited transport measurements, we identify distinct thickness regimes. CCPS flakes thinner than ~ 40 nm exhibit robust, reversible ferroelectric (FE) switching. Conversely, thicker samples display suppressed switchable polarization, with electrical conduction dominated by Cu+ migration, filamentary memristive switching, and rectification reversal. Optical illumination further enhances ionic mobility via photogenerated carriers, driving photo-assisted ionic transport in non-FE flakes and a coupled photoferroionic response in FE flakes. These findings clarify how ferroelectricity fundamentally alters ionic transport dynamics in vdW ferroionic systems.