arXiv subjects

Yuhang Tian

Publications and source records attributed to Yuhang Tian.

14 recordsLinked to original sources

arXiv2026

Med-CRAFT: An Information System for Explainable and Configurable Construction of Multimodal Medical QA Datasets

Data-intensive artificial intelligence applications increasingly rely on large-scale, high-quality, explainable, and reproducible datasets, yet the construction of such datasets often remains labor-intensive, weakly traceable, and difficult to configure. This problem is particularly critical in multimodal medical scenarios, where each question-answer sample should be semantically consistent, grounded in visual and temporal evidence, and controllable in terms of reasoning complexity. To address these challenges, we propose Med-CRAFT, an information system for explainable and configurable construction of multimodal medical question answering datasets from instructional videos. Med-CRAFT organizes dataset construction as a provenance-aware pipeline that transforms raw medical instructional videos into structured operation knowledge graphs, evidence-grounded reasoning paths, and natural-language question-answer pairs.

cs.AI↗

arXiv2026

How Much, Then Where: Credit-Conserving Action-to-Token Allocation for Multi-Turn Agent Reinforcement Learning

Credit assignment in multi-turn agent reinforcement learning operates at two levels: assigning trajectory-level credit to actions and distributing each action's credit across its tokens. In this paper, we introduce FACTOR, which separates these decisions. FACTOR uses checkpoint-calibrated TD residuals to assign per-action credits that telescope to the trajectory advantage, and feedback-conditioned teacher-student likelihood gaps to allocate each credit across the realized action tokens. Per-action normalization preserves the action-average coefficient and prevents token-level sign flips. We pair this construction with an action-mean reduction, removing the implicit dependence of an action's scalar surrogate weight on its token length. At the behavior policy and before clipping, each action's inner action-mean surrogate equals its TD credit. FACTOR consistently improves over competitive baselines across ALFWorld, WebShop, and ScienceWorld, with every environment-seed comparison favoring FACTOR and the largest gains emerging on the longest-horizon environment. The same hyperparameters transfer without retuning to a larger backbone and to a different model family. Ablations identify TD action credit as the dominant driver of the improvement, with hindsight token allocation contributing complementary gains.

cs.AI↗

arXiv2026

HarnessCompass: Guiding Automatic Harness Evolution toward Generalizable and Effective Agent Harnesses

Harness design plays a critical role in agent performance by shaping how large language models (LLMs) perceive, reason over, and act within executable environments. Recent work has proposed automatic harness evolution, which iteratively improves the harness from agent--environment interactions. However, existing methods often overfit to the evolution tasks, rely exclusively on trajectory-derived signals, and optimize harness components jointly, causing interference across components. We propose HarnessCompass, a novel automatic harness evolution framework built around constrained evolution, proactive feedback, and component-wise optimization. HarnessCompass first enforces global constraints on evolution, restricting modifications to task-agnostic harness changes that generalize beyond the evolution tasks. It then augments trajectory-derived evidence with proactive first-person feedback from the agent about harness usage, yielding richer signals for evolution. Finally, it decouples the optimization of different harness components before consolidating them into a unified harness, reducing cross-component interference while preserving component synergy. On SWE-bench Verified with GPT-5.4, HarnessCompass improves Pass@1 from 54\% to 66\% in only 5 evolution iterations, outperforming AHE in both effectiveness and evolution efficiency. In addition, the evolved harness transfers effectively to held-out tasks and other models, demonstrating substantially stronger generalization than prior automatic harness evolution methods.

cs.LG↗

arXiv2026

Overview of the NLPCC 2026 Shared Task 1: Difficulty-Aware Multilingual and Multimodal Medical Instructional Video Understanding Evaluation

Following the CMIVQA, MMI-VQA, and M4IVQA challenges in NLPCC 2023--2025, we introduce the Difficulty-Aware Medical Instructional Video Question Answering (DA-MIVQA) shared task for NLPCC 2026. DA-MIVQA extends previous multilingual and multimodal medical video benchmarks by explicitly distinguishing questions according to the type and complexity of evidence required for answering. Specifically, simple questions can often be answered from subtitle-based textual cues, whereas complex questions require visual grounding, procedural understanding, and cross-modal evidence integration. The challenge contains three tracks: Difficulty-Aware Temporal Answer Grounding in Single Video (DA-TAGSV), Difficulty-Aware Video Corpus Retrieval (DA-VCR), and Difficulty-Aware Temporal Answer Grounding in Video Corpus (DA-TAGVC). The dataset is collected from public medical instructional channels, covers diverse scenarios such as first aid, emergency response, rehabilitation, nursing, and general medical education, and is manually verified with difficulty annotations. This paper presents the task motivation, dataset construction, evaluation protocol, participation overview, competition results, and representative systems of DA-MIVQA. DA-MIVQA provides a practical benchmark for evaluating medical instructional video question answering systems under varying textual, visual, temporal, and procedural reasoning requirements.

cs.CV↗

arXiv2026

PruneTIR: Inference-Time Tool Call Pruning for Effective yet Efficient Tool-Integrated Reasoning

Tool-integrated reasoning (TIR) enables large language models (LLMs) to enhance their capabilities by interacting with external tools, such as code interpreters (CI). Most recent studies focus on exploring various methods to equip LLMs with the ability to use tools. However, how to further boost the reasoning ability of already tool-capable LLMs at inference time remains underexplored. Improving reasoning at inference time requires no additional training and can help LLMs better leverage tools to solve problems. We observe that, during tool-capable LLM inference, both the number and the proportion of erroneous tool calls are negatively correlated with answer correctness. Moreover, erroneous tool calls are typically resolved successfully within a few subsequent turns. If not, LLMs often struggle to resolve such errors even with many additional turns. Building on the above observations, we propose PruneTIR, a rather effective yet efficient framework that enhances the tool-integrated reasoning at inference time. During LLM inference, PruneTIR prunes trajectories, resamples tool calls, and suspends tool usage through three components: Success-Triggered Pruning, Stuck-Triggered Pruning and Resampling, and Retry-Triggered Tool Suspension. These three components enable PruneTIR to mitigate the negative impact of erroneous tool calls and prevent LLMs from getting stuck in repeated failed resolution attempts, thereby improving overall LLM performance. Extensive experimental results demonstrate the effectiveness of PruneTIR, which significantly improves Pass@1 and efficiency while reducing the working context length for tool-capable LLMs.

cs.CL↗

arXiv2026

ActiShade: Activating Overshadowed Knowledge to Guide Multi-Hop Reasoning in Large Language Models

In multi-hop reasoning, multi-round retrieval-augmented generation (RAG) methods typically rely on LLM-generated content as the retrieval query. However, these approaches are inherently vulnerable to knowledge overshadowing - a phenomenon where critical information is overshadowed during generation. As a result, the LLM-generated content may be incomplete or inaccurate, leading to irrelevant retrieval and causing error accumulation during the iteration process. To address this challenge, we propose ActiShade, which detects and activates overshadowed knowledge to guide large language models (LLMs) in multi-hop reasoning. Specifically, ActiShade iteratively detects the overshadowed keyphrase in the given query, retrieves documents relevant to both the query and the overshadowed keyphrase, and generates a new query based on the retrieved documents to guide the next-round iteration. By supplementing the overshadowed knowledge during the formulation of next-round queries while minimizing the introduction of irrelevant noise, ActiShade reduces the error accumulation caused by knowledge overshadowing. Extensive experiments show that ActiShade outperforms existing methods across multiple datasets and LLMs.

cs.CL↗

arXiv2025

RefineCoder: Iterative Improving of Large Language Models via Adaptive Critique Refinement for Code Generation

Code generation has attracted increasing attention with the rise of Large Language Models (LLMs). Many studies have developed powerful code LLMs by synthesizing code-related instruction data and applying supervised fine-tuning. However, these methods are limited by teacher model distillation and ignore the potential of iterative refinement by self-generated code. In this paper, we propose Adaptive Critique Refinement (ACR), which enables the model to refine itself by self-generated code and external critique, rather than directly imitating the code responses of the teacher model. Concretely, ACR includes a composite scoring system with LLM-as-a-Judge to evaluate the quality of code responses and a selective critique strategy with LLM-as-a-Critic to critique self-generated low-quality code responses. We develop the RefineCoder series by iteratively applying ACR, achieving continuous performance improvement on multiple code generation benchmarks. Compared to the baselines of the same size, our proposed RefineCoder series can achieve comparable or even superior performance using less data.

cs.CL↗

arXiv2025

Reservoir-engineered squeezed lasing through the parametric coupling

We report the first experimental demonstration of squeezed lasing in a reservoir-engineered optical parametric oscillator (OPO). The OPO provides a basis of squeezed states and parametric amplification in lasing emission, whose vacuum reservoir is coupled to a squeezed vacuum generated by a second OPO. With a precisely controlled squeezing angle and strong squeezing injection, the parametric interaction in the first OPO is exponentially enhanced. It successfully circumvents the decoherence in the system, and eliminates the undesired noise of spontaneous photon emission in the OPO. As a result, the amplified parametric process simultaneously reserves the coherence and quantum properties in the first OPO, and yields a -6.1 dB squeezed laser in optical domain with a narrow linewidth and high brightness. Our work sheds light on potential applications of squeezed lasing in quantum metrology and quantum optics.

quant-ph↗

arXiv2025

M$^3$-Med: A Benchmark for Multi-lingual, Multi-modal, and Multi-hop Reasoning in Medical Instructional Video Understanding

With the rapid progress of artificial intelligence (AI) in multi-modal understanding, there is increasing potential for video comprehension technologies to support professional domains such as medical education. However, existing benchmarks suffer from two primary limitations: (1) Linguistic Singularity: they are largely confined to English, neglecting the need for multilingual resources; and (2) Shallow Reasoning: their questions are often designed for surface-level information retrieval, failing to properly assess deep multi-modal integration. To address these limitations, we present M3-Med, the first benchmark for Multi-lingual, Multi-modal, and Multi-hop reasoning in Medical instructional video understanding. M3-Med consists of medical questions paired with corresponding video segments, annotated by a team of medical experts. A key innovation of M3-Med is its multi-hop reasoning task, which requires a model to first locate a key entity in the text, then find corresponding visual evidence in the video, and finally synthesize information across both modalities to derive the answer. This design moves beyond simple text matching and poses a substantial challenge to a model's deep cross-modal understanding capabilities. We define two tasks: Temporal Answer Grounding in Single Video (TAGSV) and Temporal Answer Grounding in Video Corpus (TAGVC). We evaluated several state-of-the-art models and Large Language Models (LLMs) on M3-Med. The results reveal a significant performance gap between all models and human experts, especially on the complex multi-hop questions where model performance drops sharply. M3-Med effectively highlights the current limitations of AI models in deep cross-modal reasoning within specialized domains and provides a new direction for future research.

cs.CV↗

arXiv2025

Overview of the NLPCC 2025 Shared Task 4: Multi-modal, Multilingual, and Multi-hop Medical Instructional Video Question Answering Challenge

Following the successful hosts of the 1-st (NLPCC 2023 Foshan) CMIVQA and the 2-rd (NLPCC 2024 Hangzhou) MMIVQA challenges, this year, a new task has been introduced to further advance research in multi-modal, multilingual, and multi-hop medical instructional question answering (M4IVQA) systems, with a specific focus on medical instructional videos. The M4IVQA challenge focuses on evaluating models that integrate information from medical instructional videos, understand multiple languages, and answer multi-hop questions requiring reasoning over various modalities. This task consists of three tracks: multi-modal, multilingual, and multi-hop Temporal Answer Grounding in Single Video (M4TAGSV), multi-modal, multilingual, and multi-hop Video Corpus Retrieval (M4VCR) and multi-modal, multilingual, and multi-hop Temporal Answer Grounding in Video Corpus (M4TAGVC). Participants in M4IVQA are expected to develop algorithms capable of processing both video and text data, understanding multilingual queries, and providing relevant answers to multi-hop medical questions. We believe the newly introduced M4IVQA challenge will drive innovations in multimodal reasoning systems for healthcare scenarios, ultimately contributing to smarter emergency response systems and more effective medical education platforms in multilingual communities. Our official website is https://cmivqa.github.io/

cs.CV↗

arXiv2024

A Comprehensive Evaluation of Large Language Models on Aspect-Based Sentiment Analysis

Recently, Large Language Models (LLMs) have garnered increasing attention in the field of natural language processing, revolutionizing numerous downstream tasks with powerful reasoning and generation abilities. For example, In-Context Learning (ICL) introduces a fine-tuning-free paradigm, allowing out-of-the-box LLMs to execute downstream tasks by analogy learning without any fine-tuning. Besides, in a fine-tuning-dependent paradigm where substantial training data exists, Parameter-Efficient Fine-Tuning (PEFT), as the cost-effective methods, enable LLMs to achieve excellent performance comparable to full fine-tuning. However, these fascinating techniques employed by LLMs have not been fully exploited in the ABSA field. Previous works probe LLMs in ABSA by merely using randomly selected input-output pairs as demonstrations in ICL, resulting in an incomplete and superficial evaluation. In this paper, we shed light on a comprehensive evaluation of LLMs in the ABSA field, involving 13 datasets, 8 ABSA subtasks, and 6 LLMs. Specifically, we design a unified task formulation to unify ``multiple LLMs for multiple ABSA subtasks in multiple paradigms.'' For the fine-tuning-dependent paradigm, we efficiently fine-tune LLMs using instruction-based multi-task learning. For the fine-tuning-free paradigm, we propose 3 demonstration selection strategies to stimulate the few-shot abilities of LLMs. Our extensive experiments demonstrate that LLMs achieve a new state-of-the-art performance compared to fine-tuned Small Language Models (SLMs) in the fine-tuning-dependent paradigm. More importantly, in the fine-tuning-free paradigm where SLMs are ineffective, LLMs with ICL still showcase impressive potential and even compete with fine-tuned SLMs on some ABSA subtasks.

cs.CL↗

arXiv2022

FLVoogd: Robust And Privacy Preserving Federated Learning

In this work, we propose FLVoogd, an updated federated learning method in which servers and clients collaboratively eliminate Byzantine attacks while preserving privacy. In particular, servers use automatic Density-based Spatial Clustering of Applications with Noise (DBSCAN) combined with S2PC to cluster the benign majority without acquiring sensitive personal information. Meanwhile, clients build dual models and perform test-based distance controlling to adjust their local models toward the global one to achieve personalizing. Our framework is automatic and adaptive that servers/clients don't need to tune the parameters during the training. In addition, our framework leverages Secure Multi-party Computation (SMPC) operations, including multiplications, additions, and comparison, where costly operations, like division and square root, are not required. Evaluations are carried out on some conventional datasets from the image classification field. The result shows that FLVoogd can effectively reject malicious uploads in most scenarios; meanwhile, it avoids data leakage from the server-side.

cs.CR↗

arXiv2022

Deterministic and universal quantum squeezing gate in virtue of teleportation-like protocol

Squeezing transformation as an essential component, gives rise to the possibility to perform various tasks of quantum information processing. However, the reported squeezing gate with best performance so far is a conditional realization at the expense of low success probability, while the performance of the deterministic ones is currently circumscribed by the limited squeezing degree of the non-classical ancilla. To address this issue, we develop and demonstrate a new scheme of deterministic and universal quantum squeezing gate with the property of non-local operation attributed to the teleportation-like protocol. We demonstrate high fidelity squeezing operation, even when the level of target squeezing being up to 10 dB, where a squeezed state with non-classical noise reduction of 6.5 dB is directly observed. Moreover, we perform a high-fidelity complex operation including a Fourier transformation and a phase squeezing gate, exploring the potential of implementing the complex task of quantum information processing with the presented functional unit. Our method can be applied to distributed quantum processor, the creation of exotic non-classical states, and quantum error correction.

quant-ph↗

arXiv2020

Resource reduction for simultaneous generation of two types of continuous variable nonclassical states

We demonstrate experimentally the simultaneous generation and detection of two types of continuous variable nonclassical states from one type-0 phase-matching optical parametric amplification (OPA) and subsequent two ring filter cavities (RFCs). The output field of the OPA includes the baseband ω0 and sideband modes ω0+/-nωf subjects to the cavity resonance condition, which are separated by two cascaded RFCs. The first RFC resonates with half the pump wavelength ω0 and the transmitted baseband component is a squeezed state. The reflected fields of the first RFC, including the sideband modes ω0+/-ωf, are separated by the second RFC, construct Einstein-Podolsky-Rosen entangled state. All freedoms, including the filter cavities for sideband separation and relative phases for the measurements of these sidebands, are actively stabilized. The noise variance of squeezed states is 10.2 dB below the shot noise limit (SNL), the correlation variances of both quadrature amplitude-sum and quadrature phase-difference for the entanglement state are 10.0 dB below the corresponding SNL.

quant-ph↗