arXiv subjects

Qingyao Li

Publications and source records attributed to Qingyao Li.

17 recordsLinked to original sources

arXiv2026

SkillGate: Training In-Policy Skill Selection in Long-Horizon Agents

Agent frameworks increasingly package procedural knowledge as skills: instruction files an agent reads on demand, while public libraries now hold thousands of them. Which skill to read has thus become a decision the policy itself makes in the middle of an episode, yet no existing signal trains it. We show that the default remedy, outcome-rewarded RL over the candidate slate, cannot teach it, for a structural reason we identify and name selector credit starvation: under a broadcast, sequence-level advantage, the few tokens that name the chosen skill carry a vanishing share of the loss, and the credit they inherit is increasingly wrong-signed as trajectories lengthen. A correct choice is punished whenever the execution after it fails, even though the choice itself is among the most valuable decisions in the trajectory. Auditing a completed run's own training artifacts confirms all three properties, each worsening monotonically with horizon. SkillGate removes the failure by construction: it partitions the token support into two disjoint credit channels, outcome credit reaching only execution tokens, and a separate action-local advantage reaching exactly the skill-naming tokens, positive only when a trajectory's single read is the correct one. On five agentic benchmarks under a 16-candidate slate, SkillGate lifts a 9B policy from 40.8% to 53.2% trial success, well ahead of the identical budget spent on outcome reward alone, while cutting exposure to misleading candidates by two thirds and reading fewer skills.

cs.AI↗

arXiv2026

EMAN: Optimization-Driven Capacity Growth through Path Emergence in Multi-Task Learning

Existing multi-task learning methods rely on hard sharing, multiple paths or experts, adaptive sharing, and dynamic expansion. However, their capacity changes are usually constrained by predefined structures or triggered by task boundaries and conflict signals. This raises a fundamental question: can a network start from exact single-path computation and grow a new independent path only when persistent optimization evidence appears? We propose the Emergent Modular Atomic Network (EMAN), an optimization-driven framework for exposing an antisymmetric growth direction through latent relative phases without instantiating a second path, and for monitoring multiple decision signals during training to transform local optimization evidence into a structural decision. EMAN materializes two equal-capacity independent paths only after certification. EMAN adaptively allocates shared and task-specific representation capacity to accommodate varying task requirements. Extensive experiments on controlled rank settings, PASCAL-Context, and NYUv2 validate its effectiveness, achieving improved performance at a competitive computational cost.

cs.LG↗

arXiv2026

Harness-R1: Learning to Edit Executable Runtime Harnesses from Agent Failure Trajectories

Agents built around large language models continually accumulate interaction trajectories during deployment, yet their behavior typically remains fixed. Beyond updating model weights, these trajectories can improve the agent harness that constructs context, mediates tools, validates actions, and recovers execution. We introduce Harness-R1, the first method, to our knowledge, that makes failure-conditioned, lifecycle-wide editing of an existing executable runtime a learned capability. It post-trains a dedicated harness engineer with online reinforcement learning so that its edits are optimized for the realized task success they produce, rather than proposed by a fixed editor. A separate 9B engineer converts batches of target-agent failures into validated executable patches; fresh same-batch reruns of the frozen target provide outcome rewards, so training updates only the engineer. Cold-start supervised fine-tuning initializes this editing policy, which is then trained online with group-relative policy optimization. Across WebShop, ALFWorld, and DBBench, Harness-R1 raises vanilla Qwen3.5-9B success from 44.3% to 53.6% (+9.3 percentage points). After direct target-agent fine-tuning, a target-specific engineer raises the average further from 59.2% to 64.2% (+5.0 points); because these gains hold both before and after fine-tuning the target, Harness-R1 points toward co-evolving the harness engineer and the target agent.

cs.AI↗

arXiv2026

VAD: Attributing Visual Evidence for Target Reconstruction in Multimodal On-Policy Distillation

Multimodal on-policy distillation (OPD) transfers fine-grained visual knowledge by supervising student-generated trajectories with a privileged-view teacher. Yet its next-token corrections are source-mixed, combining visual signals with linguistic priors and teacher-specific effects. The key challenge is to estimate which corrections are supported by visual evidence, not merely where or how strongly to distill. We introduce Visual Attribution Distillation (VAD), a counterfactual target-reconstruction algorithm that estimates the visually attributable part of a teacher correction. At each student-generated prefix, VAD evaluates the same fixed teacher with the relevant evidence present and removed. The corresponding change in centered log-probabilities defines ut, a signed proxy for the visual evidence direction that estimates how revealing the evidence supports or refutes candidate tokens. VAD projects the original correction onto this proxy to obtain an intervention-aligned component and a proxy-unexplained residual, then reconstructs a student-anchored target from the former. During training, this reconstructed target supplies the primary supervision signal, while the privileged teacher contributes a weak regularizer. Across six fine-grained visual benchmarks at 4B and 9B scales, VAD outperforms direct privileged-view distillation and visual-advantage weighting. Token- level and controlled-target analyses show that the proxy-aligned component is enriched in task-relevant visual corrections and yields stronger target shifts, especially when evidence refutes a mistaken answer. These results support counterfactual target reconstruction as an effective alternative to source-mixed supervision.

cs.CV↗

arXiv2026

MMSkills: Towards Multimodal Skills for General Visual Agents

Reusable skills have become a core substrate for improving agent capabilities, yet most existing skill packages encode reusable behavior primarily as textual prompts, executable code, or learned routines. For visual agents, however, procedural knowledge is inherently multimodal: reuse depends not only on what operation to perform, but also on recognizing the relevant state, interpreting visual evidence of progress or failure, and deciding what to do next. We formalize this requirement as multimodal procedural knowledge and address three practical challenges: (I) what a multimodal skill package should contain; (II) where such packages can be derived from public interaction experience; and (III) how agents can consult multimodal evidence at inference time without excessive image context or over-anchoring to reference screenshots. We introduce MMSkills, a framework for representing, generating, and using reusable multimodal procedures for runtime visual decision making. Each MMSkill is a compact, state-conditioned package that couples a textual procedure with runtime state cards and multi-view keyframes. To construct these packages, we develop an agentic trajectory-to-skill Generator that transforms public non-evaluation trajectories into reusable multimodal skills through workflow grouping, procedure induction, visual grounding, and meta-skill-guided auditing. To use them, we introduce a branch-loaded multimodal skill agent: selected state cards and keyframes are inspected in a temporary branch, aligned with the live environment, and distilled into structured guidance for the main agent. Experiments across GUI and game-based visual-agent benchmarks show that MMSkills consistently improve both frontier and smaller multimodal agents, suggesting that external multimodal procedural knowledge complements model-internal priors.

cs.AI↗

arXiv2026

AdverMCTS: Combating Pseudo-Correctness in Code Generation via Adversarial Monte Carlo Tree Search

Recent advancements in Large Language Models (LLMs) have successfully employed search-based strategies to enhance code generation. However, existing methods typically rely on static, sparse public test cases for verification, leading to pseudo-correctness -- where solutions overfit the visible public tests but fail to generalize to hidden test cases. We argue that optimizing against a fixed, weak environment inherently limits robustness. To address this, we propose AdverMCTS, a novel adversarial Monte Carlo Tree Search framework that combats pseudo-correctness by coupling code search with active vulnerability discovery. AdverMCTS formulates generation as a minimax-style game between a Solver agent, which synthesizes code candidates, and an Attacker agent, which evolves to generate targeted corner test cases that exploit logical divergences in the current code pool. These discovered tests form a dynamic, progressively hostile filter that penalizes fragile reasoning. Extensive experiments demonstrate that AdverMCTS significantly outperforms state-of-the-art baselines, effectively reducing false positive rates and forcing the model to generalize beyond the initial constraints. The resources of this work are available at https://anonymous.4open.science/r/AdverMCTS_open-A255.

cs.SE↗

arXiv2026

Plan-MCTS: Plan Exploration for Action Exploitation in Web Navigation

Large Language Models (LLMs) have empowered autonomous agents to handle complex web navigation tasks. While recent studies integrate tree search to enhance long-horizon reasoning, applying these algorithms in web navigation faces two critical challenges: sparse valid paths that lead to inefficient exploration, and a noisy context that dilutes accurate state perception. To address this, we introduce Plan-MCTS, a framework that reformulates web navigation by shifting exploration to a semantic Plan Space. By decoupling strategic planning from execution grounding, it transforms sparse action space into a Dense Plan Tree for efficient exploration, and distills noisy contexts into an Abstracted Semantic History for precise state awareness. To ensure efficiency and robustness, Plan-MCTS incorporates a Dual-Gating Reward to strictly validate both physical executability and strategic alignment and Structural Refinement for on-policy repair of failed subplans. Extensive experiments on WebArena demonstrate that Plan-MCTS achieves state-of-the-art performance, surpassing current approaches with higher task effectiveness and search efficiency.

cs.AI↗

arXiv2025

ATGen: Adversarial Reinforcement Learning for Test Case Generation

Large Language Models (LLMs) excel at code generation, yet their outputs often contain subtle bugs, for which effective test cases are a critical bottleneck. Existing test generation methods, whether based on prompting or supervised fine-tuning, rely on static datasets. This imposes a ``fixed-difficulty ceiling'', fundamentally limiting their ability to uncover novel or more complex bugs beyond their training scope. To overcome this, we introduce ATGen, a framework that trains a test case generator via adversarial reinforcement learning. ATGen pits a test generator against an adversarial code generator that continuously crafts harder bugs to evade the current policy. This dynamic loop creates a curriculum of increasing difficulty challenging current policy. The test generator is optimized via Reinforcement Learning (RL) to jointly maximize ``Output Accuracy'' and ``Attack Success'', enabling it to learn a progressively stronger policy that breaks the fixed-difficulty ceiling of static training. Extensive experiments demonstrate that ATGen significantly outperforms state-of-the-art baselines. We further validate its practical utility, showing it serves as both a more effective filter for Best-of-N inference and a higher-quality reward source for training code generation models. Our work establishes a new, dynamic paradigm for improving the reliability of LLM-generated code.

cs.SE↗

arXiv2025

DebugTA: An LLM-Based Agent for Simplifying Debugging and Teaching in Programming Education

In programming education, Debugging and Teaching (DT) task is a common scenario where students receive assistance in correcting their erroneous code. The task involves multiple inputs, including erroneous code, error messages, reference solutions, and the question description, with the goal of generating modification suggestions to the erroneous code. However, two key challenges hinder the effectiveness of existing approaches. Firstly, the complexity and heterogeneity of inputs inherent in DT tasks significantly elevate the reasoning challenges faced by LLMs. Second, existing approaches often fail to fully leverage the availability of standard code in DT tasks, forcing models to rely solely on complex multi-step reasoning, which limits the potential of LLMs in addressing DT tasks effectively. To address these challenges, we propose DebugTA, a novel LLM-based debugging and teaching agent with specialized tools for standard code retrieval, variable substitution to align reference code, and an external compiler for real-time code analysis. Guided by explicit pedagogical and debugging principles, DebugTA acts as an agent that decomposes a complex task into sequential LLM interactions, each utilizing distinct tools for specific subtasks, thereby simplifying the logical reasoning at each step and reducing overall reasoning complexity. Furthermore, DebugTA utilizes tool calls to align the standard code with the erroneous code as much as possible, allowing the LLM to focus on logic errors within the erroneous code and improving the accuracy of the generated suggestions. To rigorously assess the quality of modification suggestions, we introduce a student simulator-teacher interaction paradigm. Experimental results on three real-world code datasets demonstrate that DebugTA consistently improves teaching effectiveness while significantly reducing computational costs.

cs.SE↗

arXiv2025

NL-Debugging: Exploiting Natural Language as an Intermediate Representation for Code Debugging

Debugging is a critical aspect of LLM's coding ability. Early debugging efforts primarily focused on code-level analysis, which often falls short when addressing complex programming errors that require a deeper understanding of algorithmic logic. Recent advancements in large language models (LLMs) have shifted attention toward leveraging natural language reasoning to enhance code-related tasks. However, two fundamental questions remain unanswered: What type of natural language format is most effective for debugging tasks? And what specific benefits does natural language reasoning bring to the debugging process? In this paper, we introduce NL-DEBUGGING, a novel framework that employs natural language as an intermediate representation to improve code debugging. By debugging at a natural language level, we demonstrate that NL-DEBUGGING outperforms traditional debugging methods and enables a broader modification space through direct refinement guided by execution feedback. Our findings highlight the potential of natural language reasoning to advance automated code debugging and address complex programming challenges.

cs.CL↗

arXiv2025

LLM4CD: Leveraging Large Language Models for Open-World Knowledge Augmented Cognitive Diagnosis

Cognitive diagnosis (CD) plays a crucial role in intelligent education, evaluating students' comprehension of knowledge concepts based on their test histories. However, current CD methods often model students, exercises, and knowledge concepts solely on their ID relationships, neglecting the abundant semantic relationships present within educational data space. Furthermore, contemporary intelligent tutoring systems (ITS) frequently involve the addition of new students and exercises, a situation that ID-based methods find challenging to manage effectively. The advent of large language models (LLMs) offers the potential for overcoming this challenge with open-world knowledge. In this paper, we propose LLM4CD, which Leverages Large Language Models for Open-World Knowledge Augmented Cognitive Diagnosis. Our method utilizes the open-world knowledge of LLMs to construct cognitively expressive textual representations, which are then encoded to introduce rich semantic information into the CD task. Additionally, we propose an innovative bi-level encoder framework that models students' test histories through two levels of encoders: a macro-level cognitive text encoder and a micro-level knowledge state encoder. This approach substitutes traditional ID embeddings with semantic representations, enabling the model to accommodate new students and exercises with open-world knowledge and address the cold-start problem. Extensive experimental results demonstrate that our proposed method consistently outperforms previous CD models on multiple real-world datasets, validating the effectiveness of leveraging LLMs to introduce rich semantic information into the CD task.

cs.CL↗

arXiv2024

RethinkMCTS: Refining Erroneous Thoughts in Monte Carlo Tree Search for Code Generation

Tree search methods have demonstrated impressive performance in code generation. Previous methods combine tree search with reflection that summarizes past mistakes to achieve iterative improvement. However, these methods face significant challenges. First, they search directly within the code language space, neglecting the underlying reasoning process critical for effective code generation. Second, reflection-based approaches merely accumulate historical errors in memory without providing correct reasoning pathways, making it difficult for subsequent search iterations to identify optimal solutions, resulting in decreased search quality. In this work, we propose RethinkMCTS, a framework that systematically explores and refines the reasoning process for code generation. Specifically, we employ MCTS to search for thoughts before code generation and integrate MCTS with a refinement mechanism called rethink, which incorporates fine-grained code execution feedback to refine erroneous thoughts during the search. It ensures the search path aligns with better reasoning, improving overall search quality. Through extensive experiments, we demonstrate that RethinkMCTS outperforms previous search-based and feedback-enhanced code generation baselines.

cs.SE↗

arXiv2024

Learning Structure and Knowledge Aware Representation with Large Language Models for Concept Recommendation

Concept recommendation aims to suggest the next concept for learners to study based on their knowledge states and the human knowledge system. While knowledge states can be predicted using knowledge tracing models, previous approaches have not effectively integrated the human knowledge system into the process of designing these educational models. In the era of rapidly evolving Large Language Models (LLMs), many fields have begun using LLMs to generate and encode text, introducing external knowledge. However, integrating LLMs into concept recommendation presents two urgent challenges: 1) How to construct text for concepts that effectively incorporate the human knowledge system? 2) How to adapt non-smooth, anisotropic text encodings effectively for concept recommendation? In this paper, we propose a novel Structure and Knowledge Aware Representation learning framework for concept Recommendation (SKarREC). We leverage factual knowledge from LLMs as well as the precedence and succession relationships between concepts obtained from the knowledge graph to construct textual representations of concepts. Furthermore, we propose a graph-based adapter to adapt anisotropic text embeddings to the concept recommendation task. This adapter is pre-trained through contrastive learning on the knowledge graph to get a smooth and structure-aware concept representation. Then, it's fine-tuned through the recommendation task, forming a text-to-knowledge-to-recommendation adaptation pipeline, which effectively constructs a structure and knowledge-aware concept representation. Our method does a better job than previous adapters in transforming text encodings for application in concept recommendation. Extensive experiments on real-world datasets demonstrate the effectiveness of the proposed approach.

cs.IR↗

arXiv2023

Adapting Large Language Models for Education: Foundational Capabilities, Potentials, and Challenges

Online education platforms, leveraging the internet to distribute education resources, seek to provide convenient education but often fall short in real-time communication with students. They often struggle to address the diverse obstacles students encounter throughout their learning journey. Solving the problems encountered by students poses a significant challenge for traditional deep learning models, as it requires not only a broad spectrum of subject knowledge but also the ability to understand what constitutes a student's individual difficulties. It's challenging for traditional machine learning models, as they lack the capacity to comprehend students' personalized needs. Recently, the emergence of large language models (LLMs) offers the possibility for resolving this issue by comprehending individual requests. Although LLMs have been successful in various fields, creating an LLM-based education system is still challenging for the wide range of educational skills required. This paper reviews the recently emerged LLM research related to educational capabilities, including mathematics, writing, programming, reasoning, and knowledge-based question answering, with the aim to explore their potential in constructing the next-generation intelligent education system. Specifically, for each capability, we focus on investigating two aspects. Firstly, we examine the current state of LLMs regarding this capability: how advanced they have become, whether they surpass human abilities, and what deficiencies might exist. Secondly, we evaluate whether the development methods for LLMs in this area are generalizable, that is, whether these methods can be applied to construct a comprehensive educational supermodel with strengths across various capabilities, rather than being effective in only a singular aspect.

cs.AI↗

arXiv2022

Symbol Rate and Carries Estimation in OFDM Framework: A high Accuracy Technique under Low SNR

Under a low Signal-to-Noise Ratio (SNR), the Orthogonal Frequency-Division Multiplexing (OFDM) signal symbol rate is limited. Existing carrier number estimation algorithms lack adequate methods to deal with low SNR. This paper proposes an algorithm with a low error rate under low SNR by correlating the signal and applying a Fast Fourier Transform (FFT) operation. By improving existing algorithms, we improve the performance of the OFDM carrier count algorithm. The performance of the OFDM's useful symbol time estimation algorithm is improved by estimating the number of carriers and symbol rate.

cs.IT↗

arXiv2022

KGAP: Knowledge Graph Augmented Political Perspective Detection in News Media

Identifying political perspectives in news media has become an important task due to the rapid growth of political commentary and the increasingly polarized political ideologies. Previous approaches focus on textual content and leave out the rich social and political context that is essential in the perspective detection process. To address this limitation, we propose KGAP, a political perspective detection method that incorporates external domain knowledge. Specifically, we construct a political knowledge graph to serve as domain-specific external knowledge. We then construct heterogeneous information networks to represent news documents, which jointly model news text and external knowledge. Finally, we adopt relational graph neural networks and conduct political perspective detection as graph-level classification. Extensive experiments demonstrate that our method consistently achieves the best performance on two real-world perspective detection benchmarks. Ablation studies further bear out the necessity of external knowledge and the effectiveness of our graph-based approach.

cs.CL↗

arXiv2021

I-WKNN: Fast-Speed and High-Accuracy WIFI Positioning for Intelligent Stadiums

Based on various existing wireless fingerprint location algorithms in intelligent sports venues, a high-precision and fast indoor location algorithm improved weighted k-nearest neighbor (I-WKNN) is proposed. In order to meet the complex environment of sports venues and the demand of high-speed sampling, this paper proposes an AP selection algorithm for offline and online stages. Based on the characteristics of the signal intensity distribution in intelligent venues, an asymmetric Gaussian filter algorithm is proposed. This paper introduces the application of the positioning algorithm in the intelligent stadium system, and completes the data acquisition and real-time positioning of the stadium. Compared with traditional WKNN and KNN algorithms, the I-WKNN algorithm has advantages in fingerprint positioning database processing, environmental noise adaptability, real-time positioning accuracy and positioning speed, etc. The experimental results show that the I-WKNN algorithm has obvious advantages in positioning accuracy and positioning time in a complex noise environment and has obvious application potential in a smart stadium.

cs.LG↗