Papers by Yibin Wang
Achieving Stronger Generation via Simple Contrastive Tuning (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Recent years have witnessed remarkable progress in large language models (LLMs). |
| Approach: | They propose a framework for contrastive decoding to enhance instruction-tuned models. |
| Outcome: | The proposed framework improves model performance without additional data or computational resources. |
MeepleLM: A Virtual Playtester Simulating Diverse Subjective Experiences (2026.acl-long)
Copied to clipboard
Zizhen Li, Chuanhao Li, Yibin Wang, Jianwen Sun, Yukang Feng, Jiaxin Ai, Fanrui Zhang, Mingzhu Sun, Yifei Huang, Kaipeng Zhang
| Challenge: | Recent advances in large language models have expanded the role of board games as creative co-designers . however, current systems lack the capacity to offer constructive critique grounded in the emergent user experience . |
| Approach: | They propose a large language model that internalizes persona-specific reasoning patterns to accurately simulate the subjective feedback of diverse player archetypes. |
| Outcome: | The proposed model outperforms commercial models in community alignment and critique quality. |
GeometryZero: Advancing Geometry Solving via Group Contrastive Policy Optimization (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing methods for auxiliary construction training are expensive and underperform . Existing Corresponding Author training methods lack self-correction capabilities in reasoning chains. |
| Approach: | They propose a reinforcement learning framework that rewards auxiliary construction with geometric reasoning by grouping construction rewards with a Length Reward. |
| Outcome: | Experiments on Geometry3K and MathVista show that GeometryZero outperforms baselines on auxiliary constructions. |
InMind: Evaluating LLMs in Capturing and Applying Individual Human Reasoning Styles (2025.emnlp-main)
Copied to clipboard
Zizhen Li, Chuanhao Li, Yibin Wang, Qi Chen, Diping Song, Yukang Feng, Jianwen Sun, Jiaxin Ai, Fanrui Zhang, Mingzhu Sun, Kaipeng Zhang
| Challenge: | Recent large language models (LLMs) have demonstrated strong reasoning abilities across complex mathematical and scientific domains. |
| Approach: | They propose a framework to assess whether LLMs can capture and apply personalized reasoning styles in social deduction games. |
| Outcome: | The proposed framework evaluates LLMs on the game Avalon and shows that they can capture and apply individualized reasoning styles. |
Robustness-Aware Word Embedding Improves Certified Robustness to Adversarial Word Substitutions (2023.findings-acl)
Copied to clipboard
| Challenge: | Embedding interval bound constraint is important for NLP models to be certified robust, but adversarial examples can be crafted by synonym substitutions. |
| Approach: | They propose a triplet loss to train robustness-aware word embeddings for better certified robustness. |
| Outcome: | The proposed method outperforms state-of-the-art certified defense baselines and generalizes well to unseen substitutions. |
QPruner: Probabilistic Decision Quantization for Structured Pruning in Large Language Models (2025.findings-naacl)
Copied to clipboard
| Challenge: | Structured pruning can reduce model size but results in significant accuracy degradation . quantization and pruning increase the difficulty of fine-tuning, requiring a more refined quantization scheme. |
| Approach: | They propose a structured pruning framework followed by a layer-wise mixed-precision quantization scheme to reduce model memory consumption during fine-tuning and inference. |
| Outcome: | Experiments on benchmark datasets show that QPruner outperforms existing methods in memory savings while maintaining or improving model performance. |
Meeseeks: A Feedback-Driven, Iterative Self-Correction Benchmark evaluating LLMs’ Instruction Following Capability (2026.findings-acl)
Copied to clipboard
Jiaming Wang, Yunke Zhao, Peng Ding, Jun Kuang, Yibin Shen, Zhe Tang, Yilin Jin, ZongYu Wang, Xiaoyu Li, Xuezhi Cao
| Challenge: | Existing models lack the ability to adhere to instructions, resulting in suboptimal performance. |
| Approach: | They propose an automated iterative instruction-following benchmark with integrated feedback mechanism. |
| Outcome: | The proposed benchmark identifies erroneous components in model responses and provides feedback accurately. |
Medico: Towards Hallucination Detection and Correction with Multi-source Evidence Fusion (2024.emnlp-demo)
Copied to clipboard
| Challenge: | Existing studies show that LLMs can confidently state non-existent facts rather than answering "I don't know". |
| Approach: | They propose a multi-source evidence fusion enhanced hallucination detection and correction framework that fuses evidence from multiple sources and iteratively revises the hallucinous content. |
| Outcome: | The proposed framework detects whether the generated content contains factual errors, provides the rationale behind the judgment, and iteratively revises the hallucinated content. |
Tool learning via Inference-time Scaling and Cycle Verifier (2025.findings-acl)
Copied to clipboard
| Challenge: | In inference-time scaling, Chain-of-Thought (CoT) data is scarce or even unavailable. |
| Approach: | They propose a method which establishes an inference cycle to synthesize user queries and CoT data. |
| Outcome: | The proposed method achieves a 75.4% pass rate and a 79.6% win rate using small models in StableToolBench. |
RankAdaptor: Hierarchical Rank Allocation for Efficient Fine-Tuning Pruned LLMs via Performance Model (2025.findings-naacl)
Copied to clipboard
| Challenge: | Current compression techniques entail structural pruning and a recovery phase that leverages the Low-Rank Adaptation algorithm. |
| Approach: | They propose a hierarchical rank allocation method that enables efficient fine-tuning of pruned LLMs according to layerwise specific recovery requirements. |
| Outcome: | The proposed algorithm outperforms state-of-the-art methods across pruning settings and LLM architectures with improvements ranging from 0.7% to 5.5%. |