arXiv subjects

Ruike Cao

Publications and source records attributed to Ruike Cao.

12 recordsLinked to original sources

arXiv2026

Natural Chlorine Enhances Antineutrino Detection in the Deep Ocean

Large-volume neutrino telescopes such as IceCube and KM3NeT exploit natural ice or water as detection media, but their sparse instrumentation makes MeV neutrino interactions difficult to identify individually. We show that chlorine naturally abundant in seawater provides an intrinsic neutron-capture target that enables efficient tagging of electron antineutrinos through inverse beta decay. In contrast to the 2.2 MeV $γ$-ray from neutron capture on hydrogen, capture on chlorine produces an $\sim$8.6 MeV $γ$-ray cascade on a shorter time scale, yielding a distinctive delayed-coincidence signature. Using measured seawater properties at the TRIDENT site and a detector-level simulation, we find that chlorine-assisted tagging can extend the core-collapse supernova 5$σ$ discovery reach from $\sim$50 to over 100 kpc. We validate the underlying detection principle with a salted-water experiment at the China Jinping Underground Laboratory, observing correlated neutron-capture signatures under suppressed cosmic muon background conditions. Natural seawater thus transforms the deep ocean into an intrinsically enhanced antineutrino detection medium, opening new opportunities for MeV neutrino astronomy and distributed measurements of reactor and geophysical antineutrinos.

hep-ex↗

arXiv2026

Do 3D Large Language Models Really Use 3D Evidence?

Recent 3D Large Language Models (3D-LLMs) have shown strong performance on scene question answering with extensive post-training. Yet, existing benchmarks rarely evaluate how 3D-LLMs respond to referential ambiguity, insufficient scene evidence, or false presuppositions, obscuring whether their responses are grounded in 3D evidence. In this work, we introduce UNICORNS, a benchmark for evaluating the reliability of 3D-LLM responses across four evidence conditions. UNICORNS contains 11,387 questions across 557 ScanNet scans, covering eight challenge categories alongside answerable controls. A response passes only if it satisfies the required response behavior and contains no scene claim unsupported by the available evidence. Evaluation of six open-sourced 3D-LLMs reveals a pronounced performance gap: even the strongest zero-shot model achieves 61.22% accuracy on answerable controls, but only 40.94% across the eight challenge categories. To address this gap, we further propose HORN, an obligation-aware post-training method combining group-relative task rewards, leaf-conditioned obligation rewards, and supervised control replay. Extensive experiments demonstrate that HORN reaches $75.1 \pm 1.0\%$ across the challenge categories and $77.2 \pm 0.8\%$ overall, outperforming all baselines without degrading answerable-control accuracy. Our findings enable a systematic evaluation and improvement of evidence-grounded 3D reasoning for 3D-LLMs (code: https://anonymous.4open.science/r/unicorns-6B38/).

cs.CV↗

arXiv2026

Uncertainty Gates the Gradient: Adaptive Updates for Generalized Few-Shot 3D Segmentation

In generalized few-shot 3D point cloud segmentation, support annotations are often compromised by sparsity and occlusion. Blindly adapting to such low quality samples risks learning spurious geometric patterns rather than robust novel concepts. Although 3D Vision-Language Models have been introduced to sup plement supervision, the potential of uncertainty to actively govern adaptation dynamics remains underexplored. Existing signal-level methods like loss reweight ing fail to prevent unreliable episodes from steering the optimization trajectory. To address this, we propose Uncertainty-Gated Adaptive (UGA), to the best of our knowledge the first framework that elevates uncertainty to update-level gating for generalized few-shot 3D point cloud segmentation. The core mod ule, Uncertainty-Gated Update (UG-Update), estimates episodic reliability via intra-class dispersion over merged support and pseudo-labeled novel supervision, and selectively gates residual gradient updates. This encourages the model to learn novel semantics from reliable episodes while attenuating updates during unreliable ones. Complementing this, we introduce an Uncertainty-Based Pseudo Label Filter (UB-Filter) to enforce learning from high-confidence dense signals. On ScanNet200 and ScanNet++ (1-/5-shot), UGA sets new state of the art with an average +3.08 HM IoU over the strongest prior baseline.

cs.CV↗

arXiv2026

MemCalib: Benchmarking and Optimizing Memory Use in LLM Agents

The effectiveness of agent memory ultimately depends on whether the underlying LLM gives each memory in context an appropriate degree of influence over its response. Yet this capability has remained largely overlooked. To assess this capability, we introduce MemCalib, a benchmark grounded in realistic memory-system scenarios for evaluating memory use and advancing optimization algorithms. Results on the MemCalib test set reveal that frontier open- and closed-source models struggle to use memory appropriately. They frequently over-use or under-use memory rather than matching each proposition's actual use to its target level, leading to biased, low-quality responses. Experiments with common post-training algorithms, including group relative policy optimization and on-policy self-distillation, further reveal a clear directional skew: trained models improve in one direction while deteriorating in the other. We therefore propose MemCalib-RL, an ordered bidirectional counterfactual credit-assignment algorithm that separates over- and under-use signals and localizes their credit to response tokens through exact atom ablation. Results across model families and scales (Qwen3-8B, Ministral-3-8B-Instruct, and Qwen3.5-35B-A3B) show that MemCalib-RL achieves the best overall performance while better balancing over-use and under-use, with gains generalizing beyond MemCalib in external benchmark evaluation. Further experiments support its design choices and robustness and provide insight into its training dynamics.

cs.LG↗

arXiv2026

RPMem: Learning Long-Term Recurrent Parametric Memory Across Sessions for LLM Agents

Long-running LLM agents require memory that persists and evolves across sessions. Text-based memory retrieves and reconstructs past interactions at every query, making long-horizon performance increasingly dependent on retrieval quality and contextual reasoning as histories grow. Parametric memory encodes experience directly into model computation, but existing approaches provide limited support for cross-session memory evolution. Their coupling to a specific backbone further restricts memory reuse after model replacement. We introduce RPMem, a two-stage architecture that compiles each session into a model-independent latent memory through forward computation and selectively integrates it with retained memory via a task-trained recurrent gate. The consolidated memory is then mapped to backbone-specific low-rank adaptation (LoRA) parameters, allowing the encoding capability to transfer when the backbone is replaced. Evaluation across three long-term memory benchmarks and five diverse backbones demonstrates broad generalization with near-constant update cost and memory footprint. With Qwen3-8B on PERMA, RPMem reaches 85.52%, outperforming the strongest parametric and text-based baselines by 5.32 and 12.98 percentage points, respectively. Ablations validate the complementary roles of session compilation and cross-session consolidation, while dynamics analyses reveal that the gate acquires task-specific memory integration strategies. These results establish RPMem as a lifecycle-independent parametric memory framework that maintains evolving cross-session memory that remains reusable across backbone replacements. Our implementation is available at https://github.com/Quark-Medical/rpmem/tree/main.

cs.CL↗

arXiv2026

COPE: Continual Personalization of LLMs under Sparse User Feedback via User Embeddings and Self-Evaluation

While Large Language Models (LLMs) have achieved remarkable results across various benchmarks, their alignment with normative values often results in homogenized responses that fail to address diverse user preferences. Existing training-free methods often occupy valuable context windows through prompt engineering, while training-based methods typically remain static post-training, failing to support the continual optimization required in real-world settings. To address these challenges, we propose COPE (Continual Optimization with Personalized embedding and self-Evaluation), a novel optimization framework tailored for real-world-motivated interaction settings with sparse user feedback. Our framework assigns learnable personalized embeddings to each user and synergistically integrates preference capture, self-evaluation calibration, and personalized response optimization within a single update step. A key innovation of our method is the use of self-evaluation to generate proxy rewards, enabling continuous model updates even when explicit user feedback is unavailable. Experiments show that COPE consistently outperforms strong training-free and training-based baselines under sparse feedback, and remains complementary to Retrieval-Augmented Prompting (RAP). Further analyses confirm COPE's reliable self-evaluation, meaningful preference patterns, stable general capabilities, and robustness under shifting preferences and alternative evaluators.

cs.LG↗

arXiv2026

Design and Evaluation of a PMT High-Voltage system for Deepsea Neutrino Telescope

We present the design and characterization of a Cockcroft--Walton (CW) high-voltage (HV) system developed for deep-sea neutrino telescopes. The system provides independently adjustable bias voltages for 31 three-inch photomultiplier tubes (PMTs) housed in a hybrid Digital Optical Module (hDOM). We describe the system architecture, control logic, and laboratory test procedures, and report the combined PMT--base performance in terms of baseline stability, gain uniformity, and timing accuracy under conditions designed to emulate the deep-sea environment. Baseline measurements show low and stable electronic noise. Gain calibrations based on single-photoelectron spectra demonstrate that all PMTs can be tuned to a common nominal gain and remain stable over multi-day operation. Transit-time-spread measurements yield values below 1.8~ns (FWHM), consistent with manufacturer specifications. These results indicate that the CW-based HV system provides the stability and timing precision required for deep-sea multi-PMT optical modules.

hep-ex↗

arXiv2026

ATPO: Adaptive Tree Policy Optimization for Multi-Turn Medical Dialogue

Effective information seeking in multi-turn medical dialogues is critical for accurate diagnosis, especially when dealing with incomplete information. Aligning Large Language Models (LLMs) for these interactive scenarios is challenging due to the uncertainty inherent in user-agent interactions, which we formulate as a Hierarchical Markov Decision Process (H-MDP). While conventional Reinforcement Learning (RL) methods like Group Relative Policy Optimization (GRPO) struggle with long-horizon credit assignment and Proximal Policy Optimization (PPO) suffers from unstable value estimation in this context, we propose a novel uncertainty-aware Adaptive Tree Policy Optimization (ATPO) algorithm. Our method adaptively allocates the rollout budget to states with high uncertainty, quantified by a composite metric of Bellman error and action-value variance. This strategy enables more accurate value estimation, while fostering more efficient and diverse exploration. To mitigate the high computational cost of tree-based RL, we introduce two key optimizations: an uncertainty-guided pruning mechanism to minimize the number of rollouts, and an asynchronous search architecture that leverages KV cache reuse to maximize inference throughput. Extensive experiments on three public medical dialogue benchmarks demonstrate that our algorithm significantly outperforms several strong baselines, culminating in Qwen3-8B model surpassing the much larger GPT-4o ($+0.92\%$ accuracy).

cs.LG↗

arXiv2025

A Cost Effective Optimization of the hybrid-DOM Design for TRIDENT

TRIDENT is a planned multi-cubic-kilometer deep-sea neutrino telescope to be built in the South China Sea, designed to rapidly discover high-energy astrophysical neutrino sources with sensitivity to all neutrino flavors. Achieving this at scale requires a detector design that balances performance with power, cost, and mechanical simplicity. This study presents a cost-effective optimization of TRIDENT's hybrid Digital Optical Module (hDOM) design, comparing configurations using high-quantum-efficiency (QE) 3-inch PMTs and larger 4-inch PMTs, the latter evaluated with both baseline and enhanced QE assumptions. Using full-chain detector simulations incorporating site-specific seawater optical properties and realistic backgrounds, we assess performance in all-flavor neutrino detection efficiency, directional reconstruction, and tau neutrino flavor identification from 1 TeV to 10 PeV. We find that if 4-inch PMTs can achieve QE comparable to 3-inch PMTs, their performance matches or improves upon that of the 3-inch design, while significantly reducing channel count, power consumption, and cost. These findings support the 4-inch PMT hDOM as a promising and scalable choice for TRIDENT's future instrumentation.

hep-ex↗

arXiv2025

MuonSLab: A plastic scintillator based detector for muon measurement in the deep ocean

Atmospheric muons are important probes for studying primary cosmic rays and extensive air showers. Additionally, they constitute a significant background for many underground and deep-sea neutrino experiments, such as TRopIcal DEep-sea Neutrino Telescope (TRIDENT). Understanding the muon flux at various depths in the deep sea is essential for validating TRIDENT simulations and guiding the development of optimized trigger strategies. This paper introduces a novel device based on plastic scintillalors and silicon photomultipliers (SiPMs) named MuonSLab, which is designed to measure muon flux in the deep sea and has the potential to be extended to other atmospheric muon property measurements. We discuss the design and instrumentation of MuonSLab and present results from several muon flux measurements, demonstrating its sensitivity to muon detection and its stability during operations across multiple locations.

hep-ex↗

arXiv2024

Front-end electronics development of large-area SiPM arrays for high-precision single-photon time measurement

TRopIcal DEep-sea Neutrino Telescope (TRIDENT) plans to incorporate silicon photomultipliers (SiPMs) with superior time resolution in addition to photomultiplier tubes (PMTs) into its detection units, namely hybrid Digital Optical Modules (hDOMs), to improve its angular resolution. However, the time resolution significantly degrades for large-area SiPMs due to the large detector capacitance, posing significant challenges for the readout electronics of SiPMs in hDOM. We analyzed the influences of series and parallel connections when constructing a large-area SiPM array and designed a series-parallel connection SiPM array with differential output. We also designed a high-speed pre-amplifier based on transformers (MABA-007159) and radio frequency amplifiers (BGA2803), and an analog multi-channel summing circuit based on operational amplifiers (LMH6629). We measured the single photon time resolution (SPTR) of a $4\times4$ SiPM (Hamamatsu S13360-3050PE) array ($12\times12~\mathrm{mm}^2$) of approximately 300 ps FWHM. This front-end readout design enables the large-area SiPM array to achieve high-precision single photon time measurement in one readout channel.

physics.ins-det↗

arXiv2023

Nonabelian Kinetic Mixing in a Confining Phase

Dark matter from a hidden sector with SU($N$) gauge symmetry can have a nonabelian kinetic mixing portal with the standard model. The dark photon becomes massive in the confining phase without the need for spontaneous symmetry breaking. Depending on the particle content of the dark sector, there can be two or more composite vectors that get kinetic mixing through a heavy mediator particle $X$. This provides a model of composite dark photons giving a portal for direct detection of dark baryons. Avoiding exotic charged relics requires additional couplings allowing $X$ to decay to dark quarks and standard model fields, leading to further portals between the dark matter and the standard model. We comprehensively study the constraints on such models from colliders, rare decays, direct detection, and big bang nucleosynthesis.

hep-ph↗