arXiv subjects

Haitao Hu

Publications and source records attributed to Haitao Hu.

8 recordsLinked to original sources

arXiv2026

IB-RL: Isolated Bilateral Reinforcement Learning for Strategic Dialogue Agents

Reinforcement learning (RL) has achieved strong results in improving large language models (LLMs) on tasks with stationary, verifiable rewards, such as mathematical reasoning and code execution. In these settings, the environment follows fixed rules and does not adapt strategically to the agent. Strategic dialogue differs in this respect: the environment is another agent that adapts to the policy, and success depends on the interaction between the two sides. Despite this interactive nature, current RL approaches typically train a target agent against a fixed counterpart or simulator. We find that this training paradigm encourages the policy to exploit counterpart-specific regularities rather than learn strategies that generalize across counterparts. We call this problem the static-counterpart mismatch, which we quantify directly in our experiments. To address it, we propose Isolated Bilateral Reinforcement Learning (IB-RL), in which the two roles coevolve through joint rollouts while each role optimizes its own reward through fully independent advantages, action masks, and update paths. We evaluate frozen policies against fully independent held-out counterparts in both domains. On Vehicle TeleSales, IB-RL achieves 89.6% Success@1, compared to 84.6% for the best unilateral RL baseline. On Deal-or-NoDeal, it reaches 98.4% agreement against DeepSeek V4 Pro, compared to 86.4% for the best unilateral baseline. These results indicate that jointly training both roles with strict peragent isolation produces policies that generalize more effectively to unseen counterparts.

cs.AI↗

arXiv2026

Programming strain-stiffening in soft composites via structural memory near jamming

Soft composite solids, comprising discrete inclusions embedded within a compliant matrix, are emerging candidates for engineering synthetic tissues and soft robotic materials. Current strategies for controlling their nonlinear mechanics, such as strain-stiffening, have primarily relied on the nonlinear elasticity of polymer matrices. Although direct contacts between inclusions may enhance stiffening responses at high densities, the role of the non-equilibrium and history-dependent nature of disordered contact networks in composite mechanics remains unexplored. In this work, by applying a mechanical training protocol near a shear-jamming phase boundary, we demonstrate that the structural memory encoded in contact networks drives a crossover from granular-like to biopolymer-like strain stiffening. Simulations of a coarse-grained composite model reveal that this biopolymer-like mechanical response emerges from enhanced non-affine reconfigurations of nearly-jammed contact networks. Without relying on matrix nonlinearity, we establish a design strategy that leverages non-equilibrium memory effects intrinsic to granular systems to achieve highly programmable strain-stiffening in soft composites.

cond-mat.soft↗

arXiv2026

PhageBench: Can LLMs Understand Raw Bacteriophage Genomes?

Bacteriophages, often referred to as the dark matter of the biosphere, play a critical role in regulating microbial ecosystems and in antibiotic alternatives. Thus, accurate interpretation of their genomes holds significant scientific and practical value. While general-purpose Large Language Models (LLMs) excel at understanding biological texts, their ability to directly interpret raw nucleotide sequences and perform biological reasoning remains underexplored. To address this, we introduce PhageBench, the first benchmark designed to evaluate phage genome understanding by mirroring the workflow of bioinformatics experts. The dataset contains 5,600 high-quality samples covering five core tasks across three stages: Screening, Quality Control, and Phenotype Annotation. Our evaluation of eight LLMs reveals that general-purpose reasoning models significantly outperform random baselines in phage contig identification and host prediction, demonstrating promising potential for genomic understanding. However, they exhibit significant limitations in complex reasoning tasks involving long-range dependencies and fine-grained functional localization. These findings highlight the necessity of developing next-generation models with enhanced reasoning capabilities for biological sequences.

cs.CL↗

arXiv2026

Two-stage robust bilevel optimization model for facility location considering operational service level under disruption risk

The bilevel facility location problem (BO-FLP) is one of the core optimization problems behind the design of many decentralized industrial systems, e.g., supply chain systems where a supplier constructs some critical facilities and then uses them to serve retailers in a cost-effective fashion, while retailers directly handle customers aiming to minimize the total unmet demand in a rather independent fashion. When uncertainty is considered, scenario-based stochastic approaches are commonly used, but they often become impractical due to insufficient data or an exponential number of scenarios. To address this issue, this paper adopts robust optimization and proposes a novel two-stage robust bilevel facility location model. Several structural properties are derived to improve both theoretical understanding and solution efficiency. Based on this, an enhanced column-and-constraint generation algorithm is developed for robust bilevel optimization with decision-dependent uncertainty, significantly improving exact solution capability over the standard method. Numerical results show that, compared to the centralized two-stage RO model, our model pays more attention to demand fulfillment, typically resulting in higher service efficiency and better utilization of supply capacity. Under a small-scale disruption, this new model delivers better service performance. However, under a large-scale disruption, the centralized model performs more effectively.

math.OC↗

arXiv2025

AgentSentinel: An End-to-End and Real-Time Security Defense Framework for Computer-Use Agents

Large Language Models (LLMs) have been increasingly integrated into computer-use agents, which can autonomously operate tools on a user's computer to accomplish complex tasks. However, due to the inherently unstable and unpredictable nature of LLM outputs, they may issue unintended tool commands or incorrect inputs, leading to potentially harmful operations. Unlike traditional security risks stemming from insecure user prompts, tool execution results from LLM-driven decisions introduce new and unique security challenges. These vulnerabilities span across all components of a computer-use agent. To mitigate these risks, we propose AgentSentinel, an end-to-end, real-time defense framework designed to mitigate potential security threats on a user's computer. AgentSentinel intercepts all sensitive operations within agent-related services and halts execution until a comprehensive security audit is completed. Our security auditing mechanism introduces a novel inspection process that correlates the current task context with system traces generated during task execution. To thoroughly evaluate AgentSentinel, we present BadComputerUse, a benchmark consisting of 60 diverse attack scenarios across six attack categories. The benchmark demonstrates a 87% average attack success rate on four state-of-the-art LLMs. Our evaluation shows that AgentSentinel achieves an average defense success rate of 79.6%, significantly outperforming all baseline defenses.

cs.CR↗

arXiv2024

Elasticity-Controlled Jamming Criticality in Soft Composite Solids

Soft composite solids are made of inclusions dispersed within soft matrices. They are ubiquitous in nature and form the basis of many biological tissues. In the field of materials science, synthetic soft composites are promising candidates for building various engineering devices due to their highly programmable features. However, when the volume fraction of the inclusions increases, predicting the mechanical properties of these materials poses a significant challenge for the classical theories of composite mechanics. The difficulty arises from the inherently disordered, multi-scale interactions between the inclusions and the matrix. To address this challenge, we systematically investigated the mechanics of densely filled soft elastomers containing stiff microspheres. We experimentally demonstrated how the strain-stiffening response of the soft composites is governed by the critical scalings in the vicinity of a shear-jamming transition of the included particles. The proposed criticality framework quantitatively connects the overall mechanics of a soft composite with the elasticity of the matrix and the particles, and captures the diverse mechanical responses observed across a wide range of material parameters. The findings uncover a novel design paradigm of composite mechanics that relies on engineering the jamming properties of the embedded inclusions.

cond-mat.soft↗

arXiv2024

Non-monotonic Rheology and Stress Heterogeneity in confined Granular suspensions

We systematically investigated the impact of boundary confinement on the shear-thickening rheology of dense granular suspensions. Under highly confined conditions, dense suspensions were found to exhibit size-dependent or even rarely reported non-monotonic ($S$-shaped) flow curves in steady states. By performing in-situ boundary stress microscopy measurements, we observed enhanced flow heterogeneities in confined suspensions, where concentrated high-stress domains propagated stably either along or against the shear direction. By comparing the boundary stress microscopy results with macroscopic flow responses, we revealed the connection between non-monotonic rheology and stress heterogeneity in confined suspensions. These findings suggest the possibility of controlling suspension rheology by imposing different boundary confinements.

cond-mat.soft↗

arXiv2021

The role of crosslinking density in surface stress and surface energy of soft solids

Surface stress and surface energy are two fundamental parameters that determine the surface properties of any materials. While it is commonly believed that the surface stress and surface energy of liquids are identical, the relationship between the two parameters in soft polymeric gels remains debatable. In this work, we measured the surface stress and surface energy of soft silicone gels with varying weight ratios of crosslinkers in soft wetting experiments. Above a critical density, $k_0$, the surface stress was found to increase significantly with crosslinking density while the surface energy remained unchanged. In this regime, we can estimate a non-zero surface elastic modulus that also increases with the ratio of crosslinkers. By comparing the surface mechanics of the soft gels with their bulk rheology, the surface properties near the critical density $k_0$ were found to be closely related to the underlying percolation transition of the polymer networks.

cond-mat.soft↗