Latest papers
Interleaved multimodal Chain-of-Thought (CoT) improves visual reasoning by incorporating auxiliary visual evidence into intermediate reasoning. However, existing approaches remain constrained by externally defined reasoning traces and…
The weight distribution bound gives a lower bound on the weight of eigenvectors for eigenvalues of distance-regular graphs. In this note, we study when the weight distribution bound is tight for the negative eigenvalue of collinearity…
We prove that every Diophantine quadruple $\{a,b,c,d\}$ over $\mathbb{Q}(i)[X]$ that contains at least one non-constant polynomial is regular; that is, $$(a+b-c-d)^2=4(ab+1)(cd+1).$$ This result is consistent with the corresponding results…
Red-giant asteroseismology yields precise stellar parameters, making it a powerful tool for studying stellar structure and evolution, as demonstrated by the Kepler mission. However, due to Kepler's limited field of view, it primarily…
Disaggregating the prefill and decoding stages of large language model (LLM) inference into two separate sets of nodes is widely adopted in today's LLM serving systems. However, such an architecture poses significant challenges for…
Neutral-atom quantum computers offer arbitrary connectivity enabled by atom transport. Some logical operations can then be simplified or reduced entirely to geometric rearrangements of the atoms. Minimizing the duration of these movements…
Deep learning has shown promise for automated tongue diagnosis in traditional Chinese medicine (TCM), yet the design space remains underexplored. We conducted a systematic ablation study spanning 20+ model versions under rigorous 5-fold…
Autonomous multi-agent systems (AMAS) built on large language models (LLMs), such as Hermes, increasingly rely on inference-time harnesses to coordinate reasoning and action. Constructing these harnesses requires substantial engineering…
As large language models (LLMs) are deployed as agents in high-stakes settings, such as medical and legal systems, understanding their deceptive capabilities is fundamental to safety. Controlled social deduction games provide a reproducible…
Awkward Array is a Python library for representing and processing nested, variable-length data that is widely used in high-energy physics. As HL-LHC analyses increasingly rely on accelerator hardware, efficient execution of irregular…
We present ReGenVC, an end-to-end generative video codec that compresses talking-head video to an ultra-low bitrate and decodes it in real time. The encoder reduces a source clip to a compact bitstream -- a neurally compressed first frame,…
A supersymmetric index has interior zeros ($|q|<1$) if and only if the arithmetic coefficients $\delta(\nu)$ underlying the supersymmetric zeta function grow exponentially at a rate set by the nearest zero. Each $\delta(\nu)$ follows from…
Actuator dead-zones are a common and troublesome nonlinearity in motion control: a band of commanded effort over which the plant does not respond, leaving a steady-state offset or a limit cycle. This paper proposes a data-driven…
The Bier sphere of a simplicial complex $K$ is defined as the deleted join of $K$ and its combinatorial Alexander dual. We focus on the class of Bier spheres of the skeleta of a simplex. Since these Bier spheres are known to be polytopal,…
Abell 3266 (A3266) is a dynamically active galaxy cluster embedded in a dense environment of galaxy groups and clusters at similar redshift. Data from the Spektrum Roentgen Gamma (SRG)/eROSITA all-sky survey enable the study of faint X-ray…
We show that if an Anosov flow on a 3-dimensional manifold has orientable stable and unstable foliations, then the complement of any filling periodic orbit is a hyperbolic manifold. This generalizes the known case of the complement of a…
The game of Hex is one of the most celebrated connection games in combinatorics. Although it is known that the first player always has a winning strategy, very little is understood about the complexity of optimal play. Following Campbell's…
Contemporary AI discourse attributes to language models properties they cannot bear: general intelligence as substrate-independent cognition, hallucination as cognitive failure, agency as autonomous goal-pursuit, sentience as emergent inner…
In this paper, we study static, computation-friendly, lossless compression formats for graphs, focusing on memory locality and operational efficiency of $k^2$-trees. We observe that their traditional level-wise layouts suffer from poor…
Machine learning for combinatorial optimization typically relies on neural constructors trained via reinforcement learning on large offline datasets for a fixed problem class-incurring high pretraining costs and generalizing poorly outside…