NLPCC 2026
Poster Sessions
The 15th CCF International Conference on Natural Language Processing and Chinese Computing
Poster Session 1
Wednesday, 4 November 2026 · 10:30–11:30 a.m. · Poster area@Foyer
| Poster ID | Title |
|---|---|
| 1 | Context-Emotion Collaborative Representation Enhancement for Multi-turn Dialogue Value Recognition |
| 2 | HAR-Mem: A Hierarchical Memory Framework with Adaptive Retrieval |
| 3 | EvoConfig: Collaborative Systems for Efficient Autonomous Environment Configuration |
| 4 | DialectS2S: End-to-End Speech Dialogue Modeling for Low-Resource Chinese Dialects |
| 5 | MCPO: Metamorphic-Causal Prompt Optimization for LLM-based Multi-Agent System |
| 6 | SEAM: A Schema-Guided External Memory System for LLM Agents |
| 7 | Knowledge-Boundary-Aware Temporal Character Memory for Role-Playing in Long-Form Narratives |
| 8 | MemGraph: A Narrative Memory Graph for Long-Term Role-Playing Dialogue |
| 9 | Teaching LLMs to Learn Tool Trialing and Execution through Environment Interaction |
| 10 | BVPA: Branch-Verified Preference Augmentation for Long-Horizon LLM Agents |
| 11 | PERLU: Pointer-guided Evidence Reranking and Long-term Update for Dialogue Memory |
| 12 | Role-Aligned Collaborative Reinforcement Training for Heterogeneous Multi-Role Dialogue |
| 13 | EmbodiedCoop: Goal-Driven Multi-Robot Collaboration in Interleaved Task Stream via Progress Memory |
| 14 | Temporal-State Agents: Diagnosis-Driven LLM Agents for Time Series Machine Learning Engineering |
| 15 | HEKT: Dual-level Heterogeneous Distillation for Knowledge Tracing with Large Language Models |
| 16 | AutoEDM: SHAP-Guided Iterative Feature Engineering for Educational Data Mining |
| 17 | KAM-ICD: An Abbreviation-Aware and Multi-Source Knowledge-Enhanced Method for Automatic ICD Coding |
| 18 | Modeling Multi-Mapping in TCM Prescription Recommendation via Hierarchical Prototype Memory |
| 19 | CHARGE: Chinese Healthcare Augmented Retrieval and Generation Framework |
| 20 | Rethinking Law: Reflective Experience Augmentation for Legal Judgment Prediction |
| 21 | DSCF: Dual-View Semantic Clustering and Cross-Modal Fusion for Radiology Report Generation |
| 22 | TCM-AndroLLM: Domain Adaptation and Evaluation of Large Language Models for Traditional Chinese Medicine Andrology |
| 23 | MediRec: Enhancing Chinese Medication Recommendation with Explainable Clinical Reasoning |
| 24 | IndusSafetyLLM: A Domain-Adapted Large Language Model for Industrial Safety Risk Management |
| 25 | Normal-Referenced Global-Local Differential Modeling for Chest X-Ray Report Generation |
| 26 | RuleAct-Safe: Evidence Interaction Representation for Auditable Construction Safety Visual Reasoning |
| 27 | MathHERMES: Exploring Error Diagnosis and Mathematical Knowledge Remediation with Large Language Models |
| 28 | EviFrame: An Evidence-Guided Framework for Medical Text Classification |
| 29 | Decoupled Hard-Negative Ranking for Automated Short Answer Grading |
| 30 | AdaptSocratic: Adaptive Multi-Agent Socratic Teaching with Dynamic Learner Modeling and Curriculum Grounding |
| 31 | JT-RAG: A Jointly Trained Temporal Retrieval-Augmented Generation Framework for Biomedical Hypothesis Generation |
| 32 | Beyond Correlation: LLM-Guided Causal Graph Learning for Legal Text Classification |
| 33 | PriorStock: Finance-Rule Priors and Return-Fluctuation Learning for News-Driven Stock Movement Prediction |
| 34 | Zero-shot cross-domain slot filling via Domain-aware Contrastive prompting with Post-hoc Entity Refining |
| 35 | Bidirectional End-to-End Learning of Retriever-Reader Paradigm for Entity Linking |
| 36 | Classical Chinese Named Entity Recognition with LLM-Augmented Multi-Source Character Fusion |
| 37 | SCMR: Schema-Conditioned Mention-to-Proxy Routing for Multi-Event Document-Level Event Extraction |
| 38 | Structure-Prior-Guided Path Interaction for Uncertain Knowledge Graph Reasoning |
| 39 | CoPlan: Dual-Stage Taxonomy Expansion via Fast Location and Slow Reasoning |
| 40 | A Code-style Unified Framework for Multimodal Information Extraction with Knowledge Enhancement and Self-Refinement |
| 41 | Boundary-Type Decoupled Preservation for Class-Incremental Named Entity Recognition |
| 42 | Confusion-Guided Relation Disambiguation for Financial Relation Extraction |
| 43 | TAPN: Temporal Adaptive Propagation Network for Interpretable Temporal Knowledge Graph Reasoning |
| 44 | SEA: A Structured Evidence Adapter for LLM-Augmented Dialogue Relation Extraction |
| 45 | SR-NER: A Self-Evolution Training Framework for Zero-Shot Named Entity Recognition |
| 46 | Multi-source Evidence-Guided Zero-Shot Event Temporal Relation Extraction |
| 47 | Candidate-to-Event: Social Event Detection with Residual Semantic Anchor Fusion |
| 48 | Learning Risk-Calibrated Ranking for Retrieval-Augmented Generation via Three-Tier Evidence Structure |
| 49 | DIR-TIR: Synergizing Textual Dialog and Generative Visual Feedback for Interactive Image Retrieval |
| 50 | TopoChunker: Topology-Aware Agentic Document Chunking Framework |
| 51 | Bridging Semantic Gaps in RAG through Generated Context Knowledge Fusion |
| 52 | Learning to Select: Attention-based Residual Selection for Unified Hierarchical Aggregation in Multimodal Recommendation |
| 53 | DISD: Domain-Informed Speculative Decoding for Generative Recommendation |
| 54 | VeriCite: Type-Aware Verification and Revision for Faithful Citations in Scientific Survey Generation |
| 55 | State-Aware Degeneration Detection and Recovery for Repetition Suppression in LLMs |
| 56 | GuideRAG: Planning-Guided, Intent-Conditioned Graph Navigation for Retrieval-Augmented Generation |
| 57 | ReCQR: Incorporating Conversational Query Rewriting to Improve Multimodal Image Retrieval |
| 58 | ABLS-Rec: Aligning Behavioral and Linguistic Spaces for LLM-based Sequential Recommendation |
| 59 | Entropy-Gap Guided Evidence Selection for Multi-Turn RAG |
| 60 | SLEEV: A Dual-Branch Slot-Level Event--Evidence Verification Framework for Chinese Criminal Judgment Fact Condensation |
| 61 | Lost-in-the-Middle in Long-Text Generation: Synthetic Dataset, Evaluation Framework, and Mitigation Paradigm |
| 62 | Role-aware Prediction and Structure-aware Preference Alignment for Concept DEF Generation |
Poster Session 2
Wednesday, 4 November 2026 · 15:30–16:30 p.m. · Poster area@Foyer
| Poster ID | Title |
|---|---|
| 63 | M-CREST: Dataset Construction and Two-Stage Reranking for Multilingual Personalized Timeline Summarization |
| 64 | More Than Static: Continual Instruction Learning based on Dynamic Orthogonality and Contrastive Learning |
| 65 | Teaching VLMs to Think Again: Advancing Mathematical Reasoning via Critique and Correction |
| 66 | FOCUS: Decoupling Expert Personas in LLMs to Enhance Domain Expert Capabilities |
| 67 | Teacher-Guided On-Policy Distillation for Large Language Models via Interleaved Inference |
| 68 | UCO-WRF: Uncertainty-Calibrated Dynamic OptScale and Weighted Reward Frequency Aggregation |
| 69 | Causal Evidence for Attention Head Imbalance in Modality Conflict Hallucination |
| 70 | Diagnosing Contrastive Decoding Transfer for Multi-Hop Knowledge Editing |
| 71 | Distribution-Aware Retraining-Free Visual Token Pruning for Vision-Large Language Models |
| 72 | Boosting Post-Training Quantization for Large Language Models via Learnable Grids and Residuals |
| 73 | Dynamic Curriculum-Guided Data Scheduling for Multi-Task RLHF |
| 74 | ICrossformer A Transformer for Implicit Cross-Channel Dependency Modeling in Multivariate Time Series Forecasting |
| 75 | EmoHarness: Diagnosing Candidate-Selection Headroom in Exact-Match Social Reasoning |
| 76 | Understanding Before Reasoning: Enhancing Chain-of-Thought with Iterative Summarization Pre-Prompting |
| 77 | LG-Sparse: LoRA-Guided Task Protection for Pruning Multi-Dialect LLMs |
| 78 | Boosting Post-training Quantization via Salient-aware Pairing and Residual Compensation |
| 79 | Chip-Tuning: Build Efficient Classification Models with Probing |
| 80 | Understanding the Bottleneck in Multi-Intent RAG: A Detection–Selection Decomposition |
| 81 | More Thinking Is Not Always Better: Structural and Semantic Factors of Complex Reasoning Tasks in LLMs |
| 82 | CORM: Criteria-Guided Preference Data Synthesis for Omni-Modality Reward Modeling |
| 83 | Simplicity Goes Far: Auditing Prompt Optimizers with Demonstration-Conditioned Prompt Search |
| 84 | Multi-Head Multi-Token Prediction with Cross-Head Teacher Forcing for Speculative Decoding |
| 85 | COPD: Contrastive On-Policy Distillation |
| 86 | Tone Negatives and Conditional Glyph Priors for Sample-Efficient Chinese BabyLM Pretraining |
| 87 | Local Semantic Enhancement and Knowledge-Guided Annotation Enhancement for Classical Chinese Paragraph Boundary Identification |
| 88 | WKFD: Rhetorical Device Recognition via World Knowledge Injection and Feature-Decoupled Reasoning |
| 89 | When More Knowledge Is Not Better: Prompt-Level Knowledge Augmentation for NL2SQL in General-Purpose LLM Workflows |
| 90 | Evaluating AIGC in Logographic Decoding: A Multimodal Analysis of Visual Attention and Neurological Engagement |
| 91 | Identifying Chinese Sentence Modality: What Makes Modality Difficult for NLP Systems? |
| 92 | Hidden in Plain Text: A Cue-Visibility Audit for Reader-Perceived Social Comparison |
| 93 | DTCD: Dual-Track Constrained Decoding for Constituency Parsing with Large Language Models |
| 94 | BMWSD: A Bidirectional Multidimensional Reasoning Framework for LLM-based Word Sense Disambiguation |
| 95 | From Measurement Instability to Contextual Adaptation: Investigating the Abstract Malleability of Cognitive Styles in Large Language Models |
| 96 | PEEF: Pairwise Event Compatibility Learning and Multi-View Dynamic Evidential Fusion for Script Event Prediction |
| 97 | Boosting Chinese Semantic Dependency Graph Parsing via Unified Modeling of Heterogeneous Semantic Parsing Tasks |
| 98 | FableGen: Training-Free Story Visualization via Narrative Factorization and Zero-Shot Identity Anchoring |
| 99 | Beyond the Encoder: Parameter-Efficient Whisper Encoder-Decoder Joint Representations for Cross-Lingual Audio Deepfake Detection |
| 100 | Zero-TTA: Zero-Shot Text-to-Avatar Synthesis |
| 101 | A Literal-Paraphrase Dual-Decoder Bidirectional Correction Model for Teochew Dialect Speech Recognition |
| 102 | Unified Spoken Language Understanding from Heterogeneous Data: Spoken NER and SA as a Case Study |
| 103 | Multimodal Conditioned Diffusive Time Series Forecasting |
| 104 | When to Use Context and Vision: Demand-aware Information Selection for Multimodal Dialogue Translation |
| 105 | ASR-Guided Preference Optimization for Low-Resource Thai and Lao TTS |
| 106 | Acoustic Retrieval-Guided Entity-Aware Speech Translation with GRPO |
| 107 | GARE-MoE: Mixture of Experts with Global and Routed Expert Collaboration for Multilingual Speech Translation |
| 108 | TriVerify: Sanskrit-Tibetan-Chinese Inter-Translation Methods for Low-Resource Classical Texts |
| 109 | Bridging the Modality Gap in Audio-Visual Speech Recognition: Progressively Bidirectional Cross-Modal Transformation |
| 110 | Multilingual Speech-to-Instruction Translation via Structured Reasoning |
| 111 | Select What to Learn, Rank What to Prefer: Efficient Reinforcement Learning for Literary Translation |
| 112 | Unbiased Scene Graph Generation by Type-Aware Message Passing on Heterogeneous and Dual Graphs |
| 113 | SGFE: Semantically Guided Feature Enhancement for Unbiased Scene Graph Generation |
| 114 | HiPrompt-CV: Scene Graph Generation via Hierarchical Language Prompting and Commonsense Fact Verification |
| 115 | Hybrid-MetricGAN: A Cascaded Time-Frequency Wiener Filter and Generative Adversarial Network for Speech Enhancement |
| 116 | R2-VideoAgent: Retrieval-Enhanced and RL-Aligned Agent for Long-form Video Question Answering |
| 117 | AdaRAG: Adaptive Retrieval-Augmented Generation for Multi-Hop Question Answering |
| 118 | MaintRAG: A Layout-Aware Agentic QA Framework for Maintenance Manuals |
| 119 | MECo: Multi-source Evidence Fusion and Budget-aware Compilation for Lightweight Conversational Memory Question Answering |
| 120 | AdaDRAG: Efficient Adaptive Reasoning for Debate-Augmented RAG |
| 121 | MedRoute-RAG: A Routed Text-Graph Retrieval-Augmented Framework for Hallucination Mitigation in Chinese Medical Question Answering |
| 122 | CoT-HM: Source-Aware Chain-of-Thought Hallucination Mitigation for Visual Question Answering |
| 123 | D2F-ReAG: Dynamic Decomposition and Filtering for Multi-Hop Reasoning-Augmented Generation |
Poster Session 3
Wednesday, 4 November 2026 · 17:00–18:00 p.m. · Poster area@Foyer
| Poster ID | Title |
|---|---|
| 124 | StepGuard: Hierarchical Chain-of-Thought with Local Rollback for Faithful Question Answering |
| 125 | TVQA: Table-Text Verification for Reliable Heterogeneous Question Answering |
| 126 | Better Decomposition, Free Aggregation: A Synthesizer-Folding Framework for Multilingual Multi-Hop Question Answering |
| 127 | Taming the Long Tail: A Simple Head-to-Tail Representation Transfer for Reinforcement Learning in Agricultural VQA |
| 128 | When APIs Change: Benchmarking LLM Agent Robustness under Tool Evolution |
| 129 | ConfusEval: A Chinese Language Vocabulary Understanding Evaluation Benchmark |
| 130 | MobileMem: Evaluating Long-Horizon Memory for Agents in Real-World Mobile Environments |
| 131 | Auditing Large Language Models as Educational Assistants and Judges |
| 132 | ECWC-Metaphor: An English-Chinese Word-Level Corpus for Cross-Lingual Metaphor Processing |
| 133 | Valid Citations, Weak Grounding: Disentangling Evaluation Signals in Chinese RAG |
| 134 | Overconfidence in LLM-as-a-Judge: Diagnosis and Confidence-Driven Solution |
| 135 | Bridging Modalities and Knowledge: A Diagnostic Dataset for Evaluating Multi-Hop Reasoning in Real-World Scenarios |
| 136 | CET-Bench: A Cross-Lingual End-to-End Technical-Intelligence Benchmark for Document-Based Retrieval-Augmented Generation |
| 137 | LiveSciBench: A Multi-Agent Dynamic Benchmark for Multimodal Scientific Literature Reasoning |
| 138 | SMEC: A Benchmark Dataset for Structured Math Error Chains in Multi-Step Reasoning |
| 139 | CCL-DRR: A Chinese Contract Clause–Legal Article Dataset for Dual-Role Risk Review |
| 140 | HateScope: Benchmarking Large Language Models for Long-Form Chinese Hate Speech Attribution |
| 141 | A Fine-Grained Trustworthiness Assessment of LLM Safety Benchmark Results |
| 142 | Large Language Models on Classical Chinese AMR Annotation: A Multi-Genre Empirical Analysis via Multi-Model Cross-Evaluation |
| 143 | LAEP: A Leakage-Audited Evaluation Protocol for Chinese LLM-Refined Text Detection |
| 144 | Why Do Metrics Think That? Diagnosing Error Overestimation in LLM-based Machine Translation Evaluation |
| 145 | Concept-Centric Evaluation of Grade-Adaptive Educational Science Question Answering |
| 146 | MQA-Score: A Fine-grained Dataset for Chinese Multi-Question Single-Answer Scoring |
| 147 | Beyond Edit Matching: Benchmarking Chinese Grammatical Error Correction and Explanation for Korean L2 Learners with LLM-based Evaluation |
| 148 | TCM-DiagCommit: A TCM-SD-Derived Benchmark for Evidence-Conditioned Syndrome Selection and Commitment |
| 149 | ConsisOCR: A Consistency-Driven Framework for Low-Cost Automated Document Annotation |
| 150 | Inference-Sequence Guided Chain-of-Thought for Multimodal Desire, Emotion and Sentiment Recognition |
| 151 | Exploiting Global Context and Adaptive Selection for Multimodal Sarcasm Detection |
| 152 | CCOM: Conflict-Aware Calibration in Ordinal Sentiment Space for Multimodal Sentiment Analysis |
| 153 | GDSSD: Gradient-Decoupled Spectral Self-Distillation for Incomplete Multimodal Sentiment Analysis |
| 154 | HALT: Hallucination-Augmented Lightweight Triplet Discrimination for Aspect Sentiment Triplet Extraction |
| 155 | Grounding-Aware Uncertainty Reasoning for Multimodal Aspect-Based Sentiment Analysis |
| 156 | Your Stance Depends on Mine: From Node Encoding to Parent-Child Interaction for Conversational Stance Detection |
| 157 | Co-occurring Emotion Prototype Learning for Multi-label Emotion Recognition in Conversation |
| 158 | Missing-modality Multi-modal Sentiment Analysis Using Robust Representation Generated by Variational Autoencoders |
| 159 | Acoustic Grounding and Conversational Reasoning for Audio-Text Emotion-Cause Triplet Extraction |
| 160 | First-principle Driven: Modeling Language Credibility Bias for Multimodal Affective Computing |
| 161 | Thought Purity: A Defense Method for Chain-of-Thought Attacks in Local Agent Frameworks |
| 162 | RevLeet: Jailbreak Attack on LLMs via Morphological Disguise and Logical Induction |
| 163 | TCRD: Trust Modeling and Counterfactual Intervention for Rumor Detection |
| 164 | PMAPN and WBGB: A Dual-Stream Multi-Scale Framework and Benchmark for Gender Bias User Profiling |
| 165 | MFHP:A Rumor Detection Method based on Multiple Dynamic Features Fusion |
| 166 | DualKR: Joint Veracity Detection and Forgery Recognition for Fake News Videos |
| 167 | SIDE: Efficient Single-Inference Dual-Sided Jailbreaking for Open-Source LLMs |
| 168 | Romanized Harmful-Intent Prompts as Phonetic-Preserving Cross-Script Evaluation for LLM Safety |
| 169 | DR.GAP: Mitigating Bias in Large Language Models using Gender-Aware Prompting with Decoupled Reasoning |
| 170 | From Detection to Construction: User-Profile-Guided Safe Action Spaces for LLM Alignment |
| 171 | BiaGuard: Low-Rank Activation-Space Decoupling for Inference-Time Debiasing of Large Language Models |
| 172 | How Little Poisoning Is Enough? Backdoor Vulnerabilities of Low-Resource Large Language Models |
| 173 | DP-QVAC: GDP-Guided Query/Value-Wise Adaptive Clipping for Differentially Private LoRA Fine-Tuning |
| 174 | Semi-Supervised Chinese Harmful Meme Detection via Feature Decoupling and Robust Pseudo-Labeling |
| 175 | ORCA: State-Driven Policy Retrieval for Compliance in Tool-Calling Agents |
| 176 | Predictive but Bounded: Quantitative-Linguistic Features in Neural Chinese Social-Media Depression Detection |
| 177 | EADD: Entropy-Aware Dynamic Defense Against User-Side Misinformation in LLM Input |
| 178 | Quantization-Robust LLM Unlearning via Module-Sensitive Updates |
| 179 | BotRHG: Reliability-Guided Hypergraph Learning for Social Bot Detection |
| 180 | PropPathSSM: Structure-Aware State Space Modeling over Propagation Paths for Fake News Detection |
| 181 | From Alignment to Cultivation: Learning Value Priors before Alignment |
| 182 | Balancing User Preferences and Safety Risks in Open-Ended Roleplay Preference Optimization |
| 183 | LLM-RaF: Decision-Decoupled Rationale Fusion for Social-Context Fake News Detection |
| 184 | NCR-GPO: News-Comment Relation-Guided Preference Optimization for Fake News Detection |
NCMMSC 2026 Joint Poster Session (Poster Session 4)
Thursday, 5 November 2026 · 10:00–11:00 a.m. · Poster area@Foyer
| Poster ID | Title |
|---|---|
| NCMMSC 2026 #1 | FusionFlow-SVC: A Gated Fusion-Based Framework for Robust and High-Fidelity Singing Voice Conversion |
| NCMMSC 2026 #2 | DeepASMR-NSpeech: A Fine-Grained Benchmark for Non-Speech ASMR Generation |
| NCMMSC 2026 #3 | Time-Layer Adaptive Alignment for Speaker Similarity in Flow-Matching Based Zero-Shot TTS |
| NCMMSC 2026 #4 | ST-News: A Shantou News Dataset for Cross-Dialect Speech-to-Subtitle Generation and Non-Verbatim Mapping Analysis |
| NCMMSC 2026 #5 | MDFNet: A Multi-Domain Fusion Network for EEG-Based Cross-Session Auditory Attention Decoding |
| NCMMSC 2026 #6 | Post-fusion Event-aware Time-Frequency Refinement for Audio Moment Retrieval |
| NCMMSC 2026 #7 | SURE-EVAL: A Systematic and Unified Agentic Framework for Reproducible Evaluation |
| NCMMSC 2026 #8 | MrCodec: A Lightweight Multiple Bitrate Scalar Quantization Neural Speech Codec for Real-Time Communication |
| NCMMSC 2026 #9 | Band-MixNet-R4: Low-Complexity Real-World Music Restoration with Teacher–Student Distillation and Hierarchical Routing |
| NCMMSC 2026 #10 | Multilingual TASU: Extending Text-Only Alignment to Multilingual Speech Recognition via Simulated CTC Posteriors |
| NCMMSC 2026 #11 | Asynchronous Voice Anonymization by Combining Adversarial Perturbations on Speaker Embedding and Waveform |
| NCMMSC 2026 #12 | VocalMind-Bridge: Acoustic Feature Coordination for Depression Detection from Participant Speech |
| NCMMSC 2026 #13 | CTT-ED: A Unified Morpheme-Tail Framework for Low-Resource Speech Recognition in Morphologically Rich Languages |
| NCMMSC 2026 #14 | From Implicit to Explicit: A Two-Stage Evidence-Driven Framework for LLM-based ASR Error Correction |
Evaluation Poster Session (Poster Session 5)
Thursday, 5 November 2026 · 13:30–17:30 · Auditorium
| Poster ID | Title | Authors |
|---|---|---|
| 8 | EVIL-Detect for NLPCC 2026 Shared Task 6: LLM-Generated Text Detection | Hongrui Bao, Hangyu Rong, Zhuoshang Wang, Yubing Ren, and Yanan Cao |
| 10 | DAMAR: Difficulty-Aware Adaptive Multi-Agent Retrieval for Medical Instructional Videos | Wosheng Deng, Mingyang Wu, Manasa Kumari, and Zhiping Huang |
| 11 | AEGIS: An Auditable Evidence-Graph Agent for Drawdown-Shielded Asset Allocation | Bohe Chen |
| 12 | A Training-Free System for NLPCC 2026 Task 6: Recovering Weak Signals with Multi-pass Voting | Lingling Shi, Xingrui Li, Shaozhen Liu, Ruiyu Fang, Shuangyong Song, and Yongxiang Li |
| 28 | SchemaTCM: A Schema-Guided Dual-Stream Reasoning Framework for TCM LLMs | Tongyu Wang, Yan Wang, Yifeng Xue, Shilin Pan, Zhengtao Feng, Songsheng Wang, and Tongxuan Zhang |
| 29 | What a Baby Language Model Generalizes and What It Must Memorize: The Winning System for the NLPCC 2026 Chinese BabyLM Challenge | Yiming Zhao, Yue Zhang, Shunda Chen, Yalu Liu, and Ziyi Han |
| 33 | NLPCC 2026 Task 10: Citation-Level Faithfulness Verification with DeBERTa Ensembles and Class-Wise Calibration | Yanling Li, Zirui Li, and Mingyu Wan |
| 35 | FinanceAI: Temporal Macro-Graph Reasoning and Bounded LLM Advice for Daily ETF Allocation | Jiawei Yang, Xuesheng Li, Cheng Li, and Xiu Li |
| 37 | Curriculum-Enhanced BabyDeBERTa for Natural Language Understanding in Chinese | Rui Lu, Jin Wang, Kangjian He, Dan Xu, and Xuejie Zhang |
| 48 | FIIA Attack Construction via Narrative Evidence Revision with Qwen3 Scaling Diagnostics | Yiyang Peng, Bo Zhang, Qizhao Wang, Yifeng Gou, and Gang Lv |
| 62 | Imbalance-Aware Cascades and Label-Preserving Evidence Fusion for Reliable Scientific Reporting | Zhijie Liu, Jiacheng Wang, Wenhan Bian, and Ping Jian |