arXiv subjects

Zhimin Hu

Publications and source records attributed to Zhimin Hu.

12 recordsLinked to original sources

arXiv2026

Distinct Radiobiological Responses to BNCT in SAS Oral Squamous Cell Carcinoma and MCF-7 Breast Cancer Cells

This work compared the radiobiological responses of SAS oral squamous cell carcinoma cells and MCF-7 breast cancer cells following accelerator-based boron neutron capture therapy (BNCT). Neutrons were generated by bombarding a lithium target with proton beams, followed by moderation to obtain sufficient thermal neutrons for BNCT irradiation. Boronophenylalanine (BPA) was used as the boron delivery agent. BNCT-induced biological responses were evaluated by gamma-H2AX immunofluorescence staining, cell-cycle analysis, apoptosis analysis, and clonogenic survival assays. BNCT induced marked gamma-H2AX foci formation in both cell lines, indicating DNA damage-associated responses after irradiation. The two cell lines further showed distinct post-irradiation outcomes. SAS cells exhibited stronger clonogenic suppression and prominent G2/M accumulation, whereas MCF-7 cells showed sustained G0/G1 accumulation and delayed apoptosis. These results suggest that BNCT sensitivity is determined by both boron accumulation and cell-line-specific biological characteristics. This work provides experimental evidence highlighting the importance of tumor-dependent cellular responses in understanding and optimizing BNCT efficacy.

physics.med-ph↗

arXiv2026

LLMs Are Not Good Strategists, Yet Memory-Enhanced Agency Boosts Reasoning

Strategic reasoning in Large Language Models (LLMs) within long-horizon environments is often limited by inconsistent subgoals. In these settings, finite attention resources prevent the model from maintaining strategic coherence over thousands of steps. This limitation leads to strategic drift, where localized decisions fail to sustain a coherent trajectory across reasoning. To address this, we introduce EpicStar, a framework that enables agents to learn memory as policy to tackle long-horizon reasoning. Specifically, the agent maintains a bank of successful past episodes as a heuristic alongside a working memory to track short-term environmental changes. During inference, a dynamic gating mechanism determines whether to execute a retrieved action directly or to perform new reasoning through a contextual fusion of the retrieved episodes and current working memory. Utilizing StarCraft II as the testbed, we evaluated EpicStar against diverse opponent styles. It significantly outperforms baseline methods, achieving higher win rates while consuming an order of magnitude fewer tokens, and it maintains this advantage consistently across difficulty levels and opponent strategies. Our findings provide compelling evidence that structured cross-episode memory is essential for enabling LLM agents to perform robust, long-term strategic execution in dynamic, autonomous settings.

cs.CL↗

arXiv2026

Language Models Represent and Transform Concepts with Shared Geometry

How concepts are represented in neural networks is a fundamental question in machine learning. The dominant view treats concept representations as stationary geometric objects. Yet concepts appear in context, and context transforms them. Drawing from neural population geometry, we formalize concept representations as point-cloud manifolds and contextual transformations as vector fields, and instantiate this framework in large language models. Across six model families of varying scales, we find that context moves each concept differently. The variance in these displacements is semantically organized, correlating with lexical concreteness and density. Importantly, both the concepts being transformed and this variance structure are shared across models: displacement structure transported from one model predicts held-out displacements in others significantly above chance. Together, these findings show that models share a common geometry not only in how concepts are represented, but more importantly in how context transforms them, a structure with richer organization than prior work has recognized.

cs.CL↗

arXiv2026

Failures and Successes to Learn a Core Conceptual Distinction from the Statistics of Language

Generic statements like "tigers are striped" and "cars have radios" communicate information that is, in general, true. However, while the first statement is true in principle, the second is true only statistically. People are exquisitely sensitive to this principled-vs-statistical distinction. It has been argued that this ability to distinguish between something being true by virtue of it being a category member versus being true because of mere statistical regularity, is a general property of people's conceptual machinery and cannot itself be learned. We investigate whether the distinction between principled and statistical properties can be learned from language itself. If so, it raises the possibility that language experience can bootstrap core conceptual distinctions and that it is possible to learn sophisticated causal models directly from language. We find that language models are all sensitive to statistical prevalence, but struggle with representing the principled-vs-statistical distinction controlling for prevalence. Until GPT-4, which succeeds.

cs.CL↗

arXiv2026

Simulation of complex DNA damage enhancement and biological effect validation for Proton-CAT

Proton therapy has been rapidly advancing due to its excellent conformal index, but its relatively low relative biological effect (RBE) has somewhat limited its therapeutic efficacy for certain tumors. To address this, we previously proposed a nitrogen-targeting Proton-Carbon-Alpha-Therapy (Proton-CAT) enhancement method. In this letter, we present combined multi-scale DNA damage simulations and in vitro cell experiments, further investigating the mechanism of the Proton-CAT. It has been show that $^{15}$N enrichment significantly enhances complex DNA damage induced by high linear energy transfer(LET) particles within tumor regions. Under 30\% $^{15}$N conditions, $α$ and $^{12}$C particle induced DSB++ increased by 175.19\% and 52.94\%, respectively. Furthermore, in vitro cell experiments using $^{15}$N-glutamine ($^{15}$N-Glu) as the $^{15}$N carrier indicated that high concentrations of $^{15}$N-Glu did not bring about significant cytotoxicity. Following 2 Gy irradiation, the cell viability in the 500 $μ$g/mL $^{15}$N-Glu treated group exhibited a net reduction of about 15.41\% compared to the control group.This indicates that the enhanced effect of Proton-CAT primarily stems from increased complex DNA damage. This work provides a theoretical basis and multi-scale research framework for the development of the Proton-CAT.

physics.med-ph↗

arXiv2026

Are More Tokens Rational? Inference-Time Scaling in Language Models as Adaptive Resource Rationality

Human reasoning is shaped by resource rationality -- optimizing performance under constraints. Recently, inference-time scaling has emerged as a powerful paradigm to improve the reasoning performance of Large Language Models by expanding test-time computation. Specifically, instruction-tuned (IT) models explicitly generate long reasoning steps during inference, whereas Large Reasoning Models (LRMs) are trained by reinforcement learning to discover reasoning paths that maximize accuracy. However, it remains unclear whether resource-rationality can emerge from such scaling without explicit reward related to computational costs. We introduce a Variable Attribution Task in which models infer which variables determine outcomes given candidate variables, input-output trials, and predefined logical functions. By varying the number of candidate variables and trials, we systematically manipulate task complexity. Both models exhibit a transition from brute-force to analytic strategies as complexity increases. IT models degrade on XOR and XNOR functions, whereas LRMs remain robust. These findings suggest that models can adjust their reasoning behavior in response to task complexity, even without explicit cost-based reward. It provides compelling evidence that resource rationality is an emergent property of inference-time scaling itself.

cs.CL↗

arXiv2026

The Representational Geometry of Number

A central question in cognitive science is whether conceptual representations converge onto a shared manifold to support generalization, or diverge into orthogonal subspaces to minimize task interference. While prior work has discovered evidence for both, a mechanistic account of how these properties coexist and transform across tasks remains elusive. We propose that representational sharing lies not in the concepts themselves, but in the geometric relations between them. Using number concepts as a testbed and language models as high-dimensional computational substrates, we show that number representations preserve a stable relational structure across tasks. Task-specific representations are embedded in distinct subspaces, with low-level features like magnitude and parity encoded along separable linear directions. Crucially, we find that these subspaces are largely transformable into one another via linear mappings, indicating that representations share relational structure despite being located in distinct subspaces. Together, these results provide a mechanistic lens of how language models balance the shared structure of number representation with functional flexibility. It suggests that understanding arises when task-specific transformations are applied to a shared underlying relational structure of conceptual representations.

cs.CL↗

arXiv2024

Proton-CAT: a Novel Strategy for Enhanced Proton Therapy

We present a nitrogen-targeting-Proton-Carbon-Alpha-Therapy method, abbreviated as Proton-CAT, which partially converts protons into carbon-12 and $α$ particles through nuclear reactions between protons and nitrogen-15. Monte Carlo simulations validated the effectiveness of the Proton-CAT, and the study specifically focused on the distribution of relative energy deposition. The results indicated that the presence of nitrogen-15 enhanced the maximum dose level of protons, resulting in more effective damage confined to tumor cells. Statistical analysis of secondary ions has shown that the Proton-CAT significantly increases the production efficiencies of carbon-12 and $α$ particles. Furthermore, it has been revealed that elevating the nitrogen-15 concentration significantly boosts the dose of carbon and $α$ particles within the tumor region. The present work would contribute to the future development of proton therapy.

physics.med-ph↗

arXiv2024

Effect of the ${\rm^{15}N(p,α)^{12}C}$ reaction on the kinetic energy release of water molecule fragmentation

In this work, we investigated the effect of ${\rm^{15}N(p,α)^{12}C}$ reaction produced by the collision between proton and ammonia monohydrate on the kinetic energy release (KER) of water molecule fragmentation. After the occurrence of the nuclear reaction, it was found that the charge states $q$ and the flight speeds $v$ are the main factors affecting the KER of water molecule fragmentation. With the value of $q/v$ increases, the KER distribution gets wider and the peak position changes more pronounced. The energy gained by each fragment is related to the mass of the fragment and the distance of the fragment from the nuclear reaction. In this study, the fragments with smaller masses and the distances far away from the nuclear reaction get higher energies. The fragments of water molecules getting higher energy may induce other factors affecting the radiotherapy effect, which needs more detailed investigations in the future.

physics.chem-ph↗

arXiv2021

Sleep Staging Based on Multi Scale Dual Attention Network

Sleep staging plays an important role on the diagnosis of sleep disorders. In general, experts classify sleep stages manually based on polysomnography (PSG), which is quite time-consuming. Meanwhile, the acquisition process of multiple signals is much complex, which can affect the subject's sleep. Therefore, the use of single-channel electroencephalogram (EEG) for automatic sleep staging has become a popular research topic. In the literature, a large number of sleep staging methods based on single-channel EEG have been proposed with promising results and achieve the preliminary automation of sleep staging. However, the performance for most of these methods in the N1 stage do not satisfy the needs of the diagnosis. In this paper, we propose a deep learning model multi scale dual attention network(MSDAN) based on raw EEG, which utilizes multi-scale convolution to extract features in different waveforms contained in the EEG signal, connects channel attention and spatial attention mechanisms in series to filter and highlight key information, and uses soft thresholding to remove redundant information. Experiments were conducted using two datasets with 5-fold cross-validation and hold-out validation method. The final average accuracy, overall accuracy, macro F1 score and Cohen's Kappa coefficient of the model reach 96.70%, 91.74%, 0.8231 and 0.8723 on the Sleep-EDF dataset, 96.14%, 90.35%, 0.7945 and 0.8284 on the Sleep-EDFx dataset. Significantly, our model performed superiorly in the N1 stage, with F1 scores of 54.41% and 52.79% on the two datasets respectively. The results show the superiority of our network over the existing methods, reaching a new state-of-the-art. In particular, the proposed method achieves excellent results in the N1 sleep stage compared to other methods.

cs.LG↗

arXiv2019

Strong Modification of Radiative Transition Rates due to a Breit-Interaction-Induced Avoided Crossing

We present the observations of x-rays emitted from the $1s2s^{2}2p_{1/2}2p_{3/2}$ inner shell excited state of B-like W and Bi ions. The relative transition rates are obtained for two dominant radiative transitions to $1s^{2}2s^{2}2p_{1/2}$ and $1s^{2}2s^{2}2p_{3/2}$. The experimental results and the comparison with rigorous relativistic calculations show that the rates of the strong electric dipole allowed $1s^22s^22p$ -- $1s2s^22p^2$ transitions are strongly modified due to a drastic change in the wavefunction caused by the Breit interaction.

physics.atom-ph↗

arXiv2016

Polarization measurement of dielectronic recombination transitions in highly charged krypton ions

We report linear polarization measurements of x rays emitted due to dielectronic recombination into highly charged krypton ions. The ions in the He-like through O-like charge states were populated in an electron beam ion trap with the electron beam energy adjusted to recombination resonances in order to produce $Kα$ x rays. The x rays were detected with a newly developed Compton polarimeter using a beryllium scattering target and 12 silicon x-ray detector diodes sampling the azimuthal distribution of the scattered x rays. The extracted degrees of linear polarization of several dielectronic recombination transitions agree with results of relativistic distorted--wave calculations. We also demonstrate a high sensitivity of the polarization to the Breit interaction, which is remarkable for a medium-$Z$ element like krypton. The experimental results can be used for polarization diagnostics of hot astrophysical and laboratory fusion plasmas.

physics.atom-ph↗