arXiv subjects

Xiao Peng

Publications and source records attributed to Xiao Peng.

9 recordsLinked to original sources

arXiv2026

APPSolver: Adaptive Patch Partitioning for Point-Wise Ship Flow Prediction on Unstructured Meshes

Large non-uniform point sets make direct attention-based surrogate modeling costly for ship hydrodynamics. We introduce APPSolver, a point-wise flow-prediction framework built around Adaptive Patch Partitioning (APP), a deterministic quadtree representation for fixed two-dimensional horizontal slices extracted from ship CFD simulations. APP assigns finer patches near the hull and coarser patches farther away, downsamples patch contents, and recovers predictions to the full reference point set. Under a corrected protocol that constructs natural $(t,t+1)$ pairs before splitting, reuses training-set normalization statistics, and reports three model seeds, learned tokenizers are more accurate than APP-Transformer, and a persistence baseline has lower one-step MAE on all three ShipBench hulls. The supported benefit of APP is therefore computational rather than universal predictive superiority: on a representative DTC input, APP-Transformer requires 1.815 GFLOPs and 1.309 ms per model forward, while a matched ablation shows that adaptive partitioning reduces MAE by 16.4-24.9\% relative to a uniform partition augmented with learned slicing. Condition encoders provide setting-dependent gains in leave-one-hull-out evaluation, but the current absolute next-state objective does not establish accurate long-horizon dynamics. These results characterize APP as a compact spatial representation with an explicit accuracy--efficiency trade-off. Code is available at https://github.com/wenhuahuo/APPSolver .

cs.AI↗

arXiv2026

Think-Augmented Function Calling: Improving LLM Parameter Accuracy Through Embedded Reasoning

Large language models (LLMs) have demonstrated remarkable capabilities in function calling for autonomous agents, yet current mechanisms lack explicit reasoning transparency during parameter generation, particularly for complex functions with interdependent parameters. While existing approaches like chain-of-thought prompting operate at the agent level, they fail to provide fine-grained reasoning guidance for individual function parameters. To address these limitations, we propose Think-Augmented Function Calling (TAFC), a novel framework that enhances function calling accuracy through explicit reasoning at both function and parameter levels. Our method introduces a universal "think" parameter augmentation that enables models to articulate their decision-making process, with dynamic optimization for parameter descriptions to improve reasoning quality. For complex parameters, TAFC automatically triggers granular reasoning based on complexity scoring, ensuring appropriate justification for critical decisions. Additionally, we propose reasoning-guided optimization to align generated reasoning with human expectations. TAFC requires no architectural modifications to existing LLMs while maintaining full API compatibility. Evaluation on ToolBench across proprietary and open-source models demonstrates significant improvements in parameter generation accuracy and reasoning coherence for multi-parameter functions, while providing enhanced interpretability for debugging AI agent behaviors.

cs.AI↗

arXiv2026

FadeMem: Biologically-Inspired Forgetting for Efficient Agent Memory

Large language models deployed as autonomous agents face critical memory limitations, lacking selective forgetting mechanisms that lead to either catastrophic forgetting at context boundaries or information overload within them. While human memory naturally balances retention and forgetting through adaptive decay processes, current AI systems employ binary retention strategies that preserve everything or lose it entirely. We propose FadeMem, a biologically-inspired agent memory architecture that incorporates active forgetting mechanisms mirroring human cognitive efficiency. FadeMem implements differential decay rates across a dual-layer memory hierarchy, where retention is governed by adaptive exponential decay functions modulated by semantic relevance, access frequency, and temporal patterns. Through LLM-guided conflict resolution and intelligent memory fusion, our system consolidates related information while allowing irrelevant details to fade. Experiments on Multi-Session Chat, LoCoMo, and LTI-Bench demonstrate superior multi-hop reasoning and retrieval with 45\% storage reduction, validating the effectiveness of biologically-inspired forgetting in agent memory systems.

cs.AI↗

arXiv2025

Capturing Unseen Spatial Heat Extremes Through Dependence-Aware Generative Modeling

Observed records of climate extremes provide an incomplete view of risk, missing "unseen" events beyond historical experience. Ignoring spatial dependence further underestimates hazards striking multiple locations simultaneously. We introduce DeepX-GAN (Dependence-Enhanced Embedding for Physical eXtremes - Generative Adversarial Network), a deep generative model that explicitly captures the spatial structure of rare extremes. Its zero-shot generalizability enables simulation of statistically plausible extremes beyond the observed record, validated against long climate model large-ensemble simulations. We define two unseen types: direct-hit extremes that affect the target and near-miss extremes that narrowly miss. These unrealized events reveal hidden risks and can either prompt proactive adaptation or reinforce a sense of false resilience. Applying DeepX-GAN to the Middle East and North Africa shows that unseen heat extremes disproportionately threaten countries with high vulnerability and low socioeconomic readiness. Future warming is projected to expand and shift these extremes, creating persistent hotspots in Northwest Africa and the Arabian Peninsula, and new hotspots in Central Africa, necessitating spatially adaptive risk planning.

cs.LG↗

arXiv2025

Design-Life Levels for Environmental Extremes: A Dependence-Aware Block-Maxima Workflow for Severity and Persistence

Environmental risk assessment often asks how large the maximum discharge, flood, or insured loss may become over a design life rather than in a single year. In environmental records, planning-horizon risk is complicated by limited record length, extremal clustering, and sub-asymptotic behavior, yet severity estimation, clustering assessment, and design-life levels are often handled separately. We develop a dependence-aware block-maxima workflow that links these tasks within a single inferential scheme. The severity branch estimates the extreme value index from sliding block-maximum quantile scaling using data-adaptive plateau selection and covariance-aware feasible generalized least squares. The persistence branch pools native block-maxima extremal-index paths over a stable block-size window to characterize extremal clustering. Design-life levels are then derived on the chosen observation clock, with the extremal index retained as a complementary descriptor of persistence. In synthetic short-record benchmarks, the main gain is improved interval calibration under overlap dependence, especially within block-maxima comparisons. Applications to Texas and Florida streamflow and National Flood Insurance Program building-payout claims show persistent hydrologic extremes but much faster escalation of insured losses across adjacent parts of the flood-risk chain. The workflow provides calibrated severity, persistence, and design-life levels for environmental design and flood-risk assessment under dependent records.

stat.ME↗

arXiv2024

Athena: Retrieval-augmented Legal Judgment Prediction with Large Language Models

Recently, large language models (LLMs) like ChatGPT, LLaMA, and Claude have prevailed in countless domains, including legal scenarios. With LLMs' rapid technological progress, the development of prompt engineering (PE) as an interface between the LLMs and real-world applications has drawn the attention of all developers. Various PE methods have been proposed to overcome real-world challenges, such as few-shot prompting, chain-of-thought, and retrieval-augmented generation (RAG). However, RAG for legal judgment prediction (LJP) is still underexplored. To address this, we propose "Athena", a novel framework cultivating RAG as a core preprocess component to enhance LLMs' performance on specialized tasks. Athena constructs a knowledge base for accusations, attached with a semantic retrieval mechanism through vectorization. Our experiments show that Athena's overall performance has improved significantly, achieving state-of-the-art results on the CAIL2018 dataset. Our ablation study on the in-context window size parameter further reproduces LLMs' "lost-in-the-middle" phenomenon with a relative positional variation. And with moderate hyper-parameter-tuning, we can achieve at most 95% of accuracy accordingly. We also study the impact of query rewriting and data distribution, providing possible directions for future research based on former analyses.

cs.CL↗

arXiv2024

Self-controller: Controlling LLMs with Multi-round Step-by-step Self-awareness

The applications of large language models (LLMs) have been widely spread across all domains. However, the basic abilities such as the controllability of LLMs are still limited. To address this, we propose "Self-controller", a novel agentic framework bringing self-awareness into LLMs' reasoning logic. The core idea of this work is to maintain states based on the LLM's response, letting the LLM become self-aware of current status and think step by step in a multi-round chain-of-thought paradigm. Our experiment on the state of textual length has shown the controllability and effectiveness of the Self-controller. We further implement a binary search algorithm to accelerate the generation process based on the linearity and monotonicity of the textual length state. Another advantage of the Self-controller comes with DeepSeek's Context Caching technology, which significantly saves computational token consumption when a cluster of conversations shares the same prefix of context. Theoretically, we prove that in this scenario the extra time complexity is $O(c \log n)$. Results of the back-of-the-envelope estimation suggest that the token consumption of our method is no more than twice as much as that of the trivial single-round generation. Furthermore, our ablation study on word constraints demonstrates the Self-controller's consistent controllability across all foundation models.

cs.CL↗

arXiv2024

Genshin: General Shield for Natural Language Processing with Large Language Models

Large language models (LLMs) like ChatGPT, Gemini, or LLaMA have been trending recently, demonstrating considerable advancement and generalizability power in countless domains. However, LLMs create an even bigger black box exacerbating opacity, with interpretability limited to few approaches. The uncertainty and opacity embedded in LLMs' nature restrict their application in high-stakes domains like financial fraud, phishing, etc. Current approaches mainly rely on traditional textual classification with posterior interpretable algorithms, suffering from attackers who may create versatile adversarial samples to break the system's defense, forcing users to make trade-offs between efficiency and robustness. To address this issue, we propose a novel cascading framework called Genshin (General Shield for Natural Language Processing with Large Language Models), utilizing LLMs as defensive one-time plug-ins. Unlike most applications of LLMs that try to transform text into something new or structural, Genshin uses LLMs to recover text to its original state. Genshin aims to combine the generalizability of the LLM, the discrimination of the median model, and the interpretability of the simple model. Our experiments on the task of sentimental analysis and spam detection have shown fatal flaws of the current median models and exhilarating results on LLMs' recovery ability, demonstrating that Genshin is both effective and efficient. In our ablation study, we unearth several intriguing observations. Utilizing the LLM defender, a tool derived from the 4th paradigm, we have reproduced BERT's 15% optimal mask rate results in the 3rd paradigm of NLP. Additionally, when employing the LLM as a potential adversarial tool, attackers are capable of executing effective attacks that are nearly semantically lossless.

cs.CL↗

arXiv2021

A Bayesian Surveillance Model to Track Variable Rainfall-Runoff Responses for Small Watersheds

Understanding dynamics of hydrological responses is essential in producing skillful runoff forecast. This can be quantitatively done by tracking changes in hydrology model parameters that represent physical characteristics. In this study, we implement a Bayesian estimation method in continuously estimating hydrology model parameters given observations of rainfall and runoff for small watersheds. The method is coupled with a conceptual hydrology model using a Gamma distribution-based Instantaneous Unit Hydrograph. The whole analytical framework is tested using synthetic data as well as observational data from the Fall Creek watershed. The results show that the Bayesian method can well track the hidden parameters that change inter-annually. Then the model is applied to examine temporal and spatial variability of the rainfall-runoff responses and we find 1) a systematic shift in the rainfall-runoff response for the Fall Creek watershed around 1943 and 2) a statistically significant relationship between rainfall-runoff responses and watershed sizes for selected NY watersheds. Our results demonstrate potential of the Bayesian estimation method as a rapid surveillance tool in monitoring and tracking changes of hydrological responses for small watersheds.

stat.AP↗