Science & Mathematics

We investigate how humanoid robots can learn a series of motor skills defined by human infant motor development scales, and we evaluate which scales are comprehensive enough to serve as self-contained guides for developing gross motor skills up to standing and walking. We represent each motor development scale as a developmental graph that guides the training of a humanoid robot. We then apply this graph within a reinforcement learning framework so that the robot learns skills following the natural phases of human motor development described by the scale. Our work demonstrates that graphs derived from three out of four popular pediatric motor scales can successfully guide humanoid robots to develop these skills. The fourth scale misses some intermediate milestones that the other scales have, which causes some failures. After adding a missing milestone in its graph, it also enables the robot to develop those skills successfully.

Curriculum learning, which organizes training samples according to their learning priority, has been widely adopted in reinforcement learning to improve training efficiency. However, existing methods commonly construct curriculum units using predefined data properties and estimate their utility by simply aggregating a single proxy signal from recent training history.Such designs may group samples with different optimization effects and fail to anticipate how their utility evolves in relation to the policy. We propose Predictive Gradient-based Bandit (PGB) Curriculum, which first discovers optimization-aware learning units by clustering projected sample-level gradients. PGB then employs a lightweight temporal predictor to forecast the future utility of each unit and converts the predictions into a Boltzmann sampling distribution for adaptive data allocation. Experiments across multiple model scales and reasoning benchmarks show that PGB consistently outperforms standard GRPO, difficulty-based curricula, and history-based scheduling strategies, demonstrating improved data utilization and training efficiency.

Large language model (LLM) coding agents, such as Claude Code, are increasingly used for security auditing, with task-specific skills providing reusable vulnerability-analysis procedures. However, their effectiveness in detecting real vulnerabilities and the reasons for their failures remain unclear. We conduct the first large-scale empirical study of 37 security skills using Claude Code with four underlying LLMs on 100 real Linux kernel vulnerabilities across ten classes, totaling 14800 runs. We evaluate vulnerability detection at the function, statement, and vulnerability-class levels. Results show that 24.4 percent of runs locate the correct function, 18.3 percent identify both the function and responsible statement, and 15.7 percent correctly classify the vulnerability. Our failure analysis reveals limitations in evidence acquisition, state-relation reasoning, code interpretation, candidate verification, investigation coverage, and reporting. Based on these findings, we propose SKILL-SIF, a self-improving framework that iteratively evolves security skills by learning from missed vulnerabilities and failure patterns. We evolve five top-performing skills using the original benchmark and evaluate them on 100 previously unseen vulnerabilities. SKILL-SIF improves fully correct vulnerability detection by up to eight times (from 2 percent to 16 percent) on unseen vulnerabilities, while aggregate accuracy across four LLMs increases by 1.85 times (from 6.05 percent to 11.20 percent). These findings highlight the limitations of existing security skills and demonstrate the potential of failure-driven skill evolution to improve vulnerability detection in coding agents.

Two-dimensional layers coupled by position-dependent interlayer tunnelling host a layer pseudo-spin whose polarization is a controllable internal degree of freedom. We introduce the quantum geometry of this layer pseudo-spin: a mixed tensor formed from the interband Berry connection and the layer matrix element, which governs the response to an oscillating in-plane electric field. Because the layer pseudo-spin is an Ising-type scalar invariant under threefold rotation, its selection rules differ from those of a pseudovector spin. In a minimal two-band model, the layer-pseudo-spin quantum metric and Zeeman--Berry curvature are nonzero, while their zone integrals vanish. An in-plane drive therefore leaves the spatially averaged polarization unchanged at linear order but imprints a moiré-periodic intracell pattern, directly imageable by layer-resolved tunnelling. For twisted MoTe$_2$ bilayer, we quantify the local dipole modulation and identify the rectified, helicity-odd, and third-harmonic channels through which a net response appears at higher order.

Shared long-term memory lets an LLM agent serving one user inadvertently retrieve another user's credentials, but it is unclear when such exposure leads to actual harm. We introduce a three-stage audit that separates three events: a foreign credential reaches the prompt (exposure), the agent acts with it (behavior), and the other user's account changes (state). In AppWorld, a store that mixes 763 past logins from all users retrieves another user's credential for 84.7 percent of required apps, yet agents rarely act on it. Misuse appears in a stress test where memory lacks the user's own credential and presents a single foreign one as the user's record: seven agents then use it in 80 to 94 percent of runs and change the victim's account in up to 43 percent, sometimes passing the official check. Naming the owner before or inside each credential line prevents most of this: with all else fixed, misuse falls from 92 to 1 of 305 paired runs, also with name-free usernames, though owner-bound notes cost 3 to 5 successful tasks per model on clean memory. A wrongly written owner largely restores misuse; trailing owner tags and added safety instructions help only some models. Agent memory should consistently record whose credential it stores, and evaluations should measure behavior, not only exposure.

Hybrid non-orthogonal multiple access (H-NOMA), which leverages the complementary advantages of orthogonal multiple access (OMA) and NOMA, has emerged as a promising technique for next-generation wireless networks. This paper investigates power allocation optimization for the OMA and NOMA time slots allocated to the opportunistic user in H-NOMA systems, aiming to overcome the limitations of conventional equal power allocation strategies. Specifically, we consider an H-NOMA scheme that integrates hybrid successive interference cancellation (HSIC) with power adaptation (PA), referred to as the HSIC-PA aided H-NOMA scheme. To better adapt to practical communication scenarios, distinct power coefficients are assigned to the OMA and NOMA time slots of the opportunistic user to maximize the achievable data rate. Closed-form expressions for the optimal power allocation coefficients are derived, and simulation results validate the accuracy and effectiveness of the theoretical analysis.

AI systems increasingly shape consequential decisions in employment, lending, healthcare, and public services, making fairness an important problem for auditing and governance. Fairness assessment often requires more than identifying disparities in outcomes; it also requires understanding how those disparities arise. Causal approaches to algorithmic fairness address this explanatory problem by representing pathways through which protected attributes and other factors may influence outcomes. Yet causal reasoning in sociotechnical systems remains uncertain: relevant confounders may be unobserved, causal directions contested, and conclusions sensitive to modeling assumptions and representational choices. We introduce causal humility as a governance framework for managing the relationship between epistemic uncertainty and causal authority. Drawing on causal fairness research, epistemology, political philosophy, and science and technology studies, causal humility has four requirements: making consequential assumptions and omissions visible; preserving contestability across defensible causal accounts; recognizing indeterminacy as a legitimate analytical outcome; and maintaining proportionality between evidentiary support and institutional authority or action. We operationalize these requirements through Causal Cards, a documentation artifact for recording the causal basis, assumptions, omissions, model dependence, and governance implications of fairness claims. Using the UCI Adult dataset as a worked epistemic demonstration, we show how different causal discovery procedures applied to the same observational data render fairness-relevant pathways visible, absent, or unresolved. Causal humility treats such disagreement as information about the limits of what can responsibly be claimed and the authority those claims should carry.

Similarity-gated coverage does not establish that retrieved evidence contains the facts needed to answer a question. We audit Mexican-Spanish financial-regulatory QA and find that 35 of 135 covered items lack some or all load-bearing facts. On the post-hoc full-evidence slice (n=98), eleven prompt-matched models score 95-98/98 and none differs significantly from the anchor; this does not establish equivalence. Removing evidence-defective items changes the nominal significance status of 14 of 55 pairwise comparisons. On 56 zero-evidence items, accuracy spans 33.9%-89.3%. Removing evidence from the audited items under an abstention-permitting prompt reduces accuracy by 63-81 percentage points in all ten tested legs, with no recovered items. Building on sufficient-context research, we contribute MexRegQA v1.0: 192 items with provenance and evidence-status labels, frozen retrieval evidence for 191 evaluated items, predictions, judge outputs and offline checks. The audit was outcome-aware, the later model/panel sweep reused the same questions, and all reported scores and labels rely on machine evaluation without independent human validation. We propose releasing evidence-status labels alongside grounded QA benchmarks so that model comparisons can be examined separately from evidence availability.

Humanoid robots operating in human-centric environments must coordinate the entire body to traverse tight, cluttered spaces while maintaining balance. Existing methods often struggle to jointly consider dynamic stability and collision avoidance from high-dimensional perceptual inputs. Moreover, for humanoids, safe behavior is inherently multimodal: the same obstacle may admit multiple valid traversals, which is difficult to represent with a single deterministic policy. To address these challenges, we propose SWAP (Steerable Whole-body Avoidance Policy), a hierarchical framework that separates collision-aware decision making from stabilization. A high-level, perception-conditioned diffusion policy generates multimodal, collision-aware motion proposals, while a low-level whole-body RL controller tracks these proposals and maintains dynamic balance. To efficiently train the high-level policy, we obtain scalable supervision via hindsight labeling: we roll out the low-level policy under randomized commands, place obstacles only in the space that each executed motion did not sweep, so the motion is collision-free by construction, and re-render the perception the robot would have seen. At deployment, Monte Carlo sampling with simple cost signals steers the diffusion model toward safe, task-aligned behaviors without heavy reward tuning. Experiments in simulation and on a real Unitree G1 humanoid show reliable avoidance with static and dynamic obstacles. Test-time steering further enables customization, improving targeted rewards (e.g., speed, orientation, or posture) by 34% on average while maintaining a high collision avoidance rate. Together, these results support a practical solution to humanoid whole-body collision avoidance with test-time steerability. Project website: https://iscoyizj.github.io/SWAP-steerable-humanoid-oa-website/

Medical data remain fragmented across institutional silos because of privacy, governance, and ownership constraints. Federated learning enables medical institutions to adapt large language models (LLMs) collaboratively without transferring raw records. However, independently parameterized low-rank adapters incur repeated parameter and communication costs across Transformer projections, while uniform aggregation may limit adaptation to heterogeneous institutional data. We propose PFedQFT, a quantum-inspired framework for personalized federated fine-tuning with a compact shared adapter. Specifically, PFedQFT constructs reusable orthonormal adaptation bases from compact skew-symmetric generators using a truncated matrix-exponential approximation followed by QR orthonormalization. Compatible projections reuse these bases, while a globally shared full core preserves the flexibility of each projection. Each client further maintains a lightweight private residual operator, whose directions are learned globally and whose coefficients remain local. Experiments on seven public medical tasks and multiple LLM backbones show substantial reductions in trainable parameters and per-round adapter upload volume, with competitive predictive performance on most tasks.

Trust and reputation mechanisms can reduce the influence of malicious clients in federated learning (FL), but strict update filtering is difficult under non-independent and identically distributed (non-IID) data because honest local updates may temporarily reduce performance on a shared validation proxy. Extending an earlier trust-based semi-decentralized FL (SDFL) framework, this paper operationalizes the trust signals and introduces peer-relative update screening that combines a robust validation-utility score with direction- and norm-based update quality. The method is evaluated with 50 MNIST clients, 20% persistent label-flip attackers, Dirichlet heterogeneity levels of 0.5 and 0.1, 100 communication rounds, and five paired random seeds. Under moderate heterogeneity, the proposed method reaches 98.37% final accuracy while rejecting 95.82% of malicious updates and retaining 90.30% of benign updates; the final benign and malicious trust distributions are completely separated (KS=1.00). Under severe heterogeneity, it reaches 95.25% accuracy, rejects 88.55% of malicious updates, and retains 76.90% of benign updates (KS=0.96). Trust weighting alone preserves slightly higher utility in the most heterogeneous condition, but accepts all malicious submissions. The results therefore distinguish influence mitigation from explicit containment and quantify the utility-selectivity trade-off introduced by screening. The study provides an empirically validated off-chain governance core that can feed blockchain-based audit and incentive layers in SDFL.

Modern generative models can produce visually plausible images while violating exact identity constraints. Evaluating such failures requires distinguishing violations of the requested identity from permissible variation in appearance. We introduce Low-Entropy Visual Identity (LEVI), which characterizes visual identities determined by compact facts and explicit rules. The low-entropy requirement concerns identity uncertainty, while appearance may remain diverse. We formulate identity fidelity in terms of the probability that a conditional image distribution assigns to the identities permitted by the prompt. Visual Identity Spaces (VIS) formalize this distinction by specifying admissible identity values and which different realizations express the same identity. Building on this formulation, we develop Verifiable Visual Grammar (VVG), a declarative language whose typed primitives and compositional constraints describe identity-bearing structures. Using this grammar, we build a benchmark for image generation. It assesses each output against the task's identity requirements, drawing evidence from the final pixels. The curated benchmark contains 1,375 tasks spanning 120 base identity spaces across six domains. Among ten evaluated model configurations, the highest whole-task pass rate is 58.69%. These results reveal a persistent gap between visually plausible images and faithful realization of their specified identities. Project page: https://maxiao0234.github.io/LEVI/.

Micro-expert pruning reduces the capacity of mixture-of-experts (MoE) models by removing coupled channels within routed experts. However, finer pruning granularity alone does not guarantee the preservation of downstream capabilities, as similar language-modeling perplexity can accompany substantially different task performance. We revisit micro-expert mask construction as two distinct decisions: allocating retained capacity across experts and selecting channels within that capacity. We propose UC-RAMP, a two-stage post-training pruning framework that explicitly separates these decisions. CAPE uses cross-domain calibration evidence to redistribute expert widths under an exact layer-wise budget, while RAMP refines channel composition at fixed widths through one-out, one-in exchanges guided by a channel-interaction surrogate. Across three MoE models at 25% and 50% routed-expert channel pruning, UC-RAMP achieves the highest ten-task mean accuracy among evaluated pruning methods in five of six settings. At 50% pruning, it outperforms the strongest evaluated baseline by 8.19, 3.96, and 6.14 percentage points across the three models, respectively. In a single-A100 deployment at 25% pruning, UC-RAMP reduces measured time to first token by 19.7--23.9% and time per output token by 20.2--22.3% relative to the unpruned model. The code is available at https://github.com/xzktx003/UC_RAMP.

Future lunar surface missions require reliable positioning for autonomous rover operations. Lunar navigation satellite systems may provide limited vertical positioning accuracy due to satellite availability and observation geometry, particularly during early deployment. This study investigates a radiometric navigation architecture in which a lunar surface station (SS) improves rover positioning through double-differenced satellite observations and additional surface-to-surface ranging. Pseudorange and Doppler measurements from lunar navigation satellites are used for relative positioning between the SS and rover, while dual one-way ranging (DOWR) provides an additional range observable. The radiometric solution is integrated with rover dead-reckoning information. Monte Carlo simulations were conducted using five satellites representative of NASA's Lunar Communications Relay and Navigation Systems (LCRNS) constellation and a rover trajectory near Connecting Ridge at the lunar south pole. The results show that the SS improves horizontal and vertical relative positioning, with further gains from DOWR. In the simulated scenario, the mean 99.7% vertical positioning error decreased from 29.9 m without the SS to 5.5 m with double-differenced pseudorange and Doppler processing, and to 4.8 m with DOWR. An uncorrected DOWR hardware bias significantly degraded positioning, whereas the proposed bias-estimation method achieved a final estimation error of 0.058 +/- 0.057 m. The benefit of DOWR also depended on observation geometry. These results demonstrate the potential of a lunar surface station to improve vertical relative positioning through direct radiometric ranging, complementing its communication and navigation support roles.

The block generalized-power method pairs threshold-based support selection with inexpensive Stiefel updates, but the standard block formulation conflates two nonequivalent objectives and leaves inactive components, threshold ties, and post-selection refitting ill-specified. We rebuild the model from the primal. Eliminating each loading column from the normalized loading-ball problem gives a sum of soft-thresholded Euclidean norms (A-SUM); the squared-threshold objective in common use arises from a quadratic lift (A-SQ). We prove both identities, exhibit a two-component instance whose maximizers disagree, and specify explicit conventions for zero-response columns, ties, rank deficiency, and mask-before-refit. For A-SQ we separate the historical full-polar iteration from a safeguarded Riemannian ascent and prove, for the latter, whole-sequence convergence via a sufficient-ascent lemma, a relative-error lemma, and the Kurdyka-Lojasiewicz property, which applies because the soft-threshold objective and the Stiefel constraint are semi-algebraic. A strict threshold margin then forces finite identification of the support and sign pattern. The post-identification map is ordinary power iteration on the active covariance for a rank-one hard threshold, and the Frechet derivative of a coupled polar map for component-specific supports; a radius-based safe-screening rule follows from the same local analysis. Planted-spike experiments match the predicted identification time and local rates. On modern-AI panels (GPT-2 residual stream, ViT CLS stream, Pitprops) the thresholded block-sparse components match dense PCA on reconstruction and improve on it in monosemanticity and class specialisation. On three native-label scRNA datasets (Zheng17 FACS-sorted PBMCs, 68k donor-A with FACS-correlated labels, Tabula Muris) A-SQ beats dense PCA on cluster-recovery AMI by +0.023 to +0.044 at matched sparsity.