13 papers
On Stable Long-Form Generation: Benchmarking and Mitigating Length Volatility
Zhitao He, Haolin Yang, Rui Min +2
Large Language Models (LLMs) excel at long-context understanding but exhibit significant limitations in long-form generation. Existing studies primarily focus on single-generation…
RebuttalAgent: Strategic Persuasion in Academic Rebuttal via Theory of Mind
Zhitao He, Zongwei Lyu, Yi R Fung
Although artificial intelligence (AI) has become deeply integrated into various stages of the research workflow and achieved remarkable advancements, academic rebuttal remains a si…
ClinTutor-R1: Advancing Scalable and Robust One-to-Many Alignment in Clinical Socratic Education
Zhitao He, Haolin Yang, Zeyu Qin +1
While Large Language Models (LLMs) have achieved remarkable success in dyadic (one-on-one) instruction, they face significant challenges in One-to-Many alignment, such as clinical…
MARS-SQL: A multi-agent reinforcement learning framework for Text-to-SQL
Haolin Yang, Jipeng Zhang, Zhitao He +2
Large Language Models (LLMs) often struggle with the precise logic and schema alignment required for complex Text-to-SQL tasks. While current methods rely heavily on static prompti…
SELF-REDRAFT: Eliciting Intrinsic Exploration-Exploitation Balance in Test-Time Scaling for Code Generation
Yixiang Chen, Tianshi Zheng, Shijue Huang +2
Test-time scaling without interpreter feedback is essential for real-world code generation scenarios where test cases are not readily available. While existing paradigms often rely…
Lean4Physics: Comprehensive Reasoning Framework for College-level Physics in Lean4
Yuxin Li, Minghao Liu, Ruida Wang +6
We present **Lean4PHYS**, a comprehensive reasoning framework for college-level physics problems in Lean4. **Lean4PHYS** includes *LeanPhysBench*, a college-level benchmark for for…
Reasoning Path Divergence: A New Metric and Curation Strategy to Unlock LLM Diverse Thinking
Feng Ju, Zeyu Qin, Rui Min +3
While Test-Time Scaling (TTS) has proven effective in improving the reasoning ability of large language models (LLMs), low diversity in model outputs often becomes a bottleneck; th…
CultureCLIP: Empowering CLIP with Cultural Awareness through Synthetic Images and Contextualized Captions
Yuchen Huang, Zhiyuan Fan, Zhitao He +3
Pretrained vision-language models (VLMs) such as CLIP excel in general multimodal comprehension but often struggle to capture nuanced, context-dependent visual cues. This makes it…
MMBoundary: Advancing MLLM Knowledge Boundary Awareness through Reasoning Step Confidence Calibration
Zhitao He, Sandeep Polisetty, Zhiyuan Fan +3
In recent years, multimodal large language models (MLLMs) have made significant progress but continue to face inherent challenges in multimodal reasoning, which requires multi-leve…
MATP-BENCH: Can MLLM Be a Good Automated Theorem Prover for Multimodal Problems?
Zhitao He, Zongwei Lyu, Dazhong Chen +2
Numerous theorems, such as those in geometry, are often presented in multimodal forms (e.g., diagrams). Humans benefit from visual reasoning in such settings, using diagrams to gai…
Mathematical Proof as a Litmus Test: Revealing Failure Modes of Advanced Large Reasoning Models
Dadi Guo, Jiayu Liu, Zhiyuan Fan +5
Large reasoning models (e.g., R1, o3) have demonstrated remarkable mathematical problem-solving abilities. However, the high reported accuracy of these advanced models on popular d…
MedEBench: Diagnosing Reliability in Text-Guided Medical Image Editing
Minghao Liu, Zhitao He, Zhiyuan Fan +2
Text-guided image editing has seen significant progress in natural image domains, but its application in medical imaging remains limited and lacks standardized evaluation framework…
MAC-Tuning: LLM Multi-Compositional Problem Reasoning with Enhanced Knowledge Boundary Awareness
Junsheng Huang, Zhitao He, Yucheng Huang +3
The hallucination of non-existent facts by LLMs is an important problem given its widespread adoption across various applications. Previous research addresses this problem by analy…