CLangSAM-2: Clinician-Language-Guided Segment Anything Model for Intracranial Hemorrhage Segmentation

Semantic Fusion for Language-Guided CT Segmentation

RuiTian

NLP AI Engineer Intern

• Built a multistage document content filter using rerank and lightweight LLM, removing redundant context while preserving 94.2% of effective content and reducing token usage by 30.6%.
• Enhanced zero-shot GRPO with entropy regularization, achieving comparable performance with only 1/4 of the rollouts.

SyRHM: Symbolic Reasoning with Retrieval for Zero-shot Harmful Meme Detection

Meme hateful content detection through Symbolic CoT

Structured Confidence Perturbation: A Geometric Paradigm for Robust Adversarial Training

information-geometric framework formulates adversarial optimization

Light Weight Deep Learning for Automated 3D Dental Finish Line Detection

Oral Scan Data Reformation and Segmentation

Pharmolix

AI Developer Intern

• Built a Milvus–FastAPI hybrid retrieval system with query rewriting and chunk-level fusion for document QA.
• Developed a PDF-grounded annotation tool enabling accurate chunk-to-document alignment.
• Constructed a LightRAG-based knowledge graph for context-aware retrieval across large corpora.

Tencent Light Creative Camp Best Creative Award

A biodiversity science popularization mini-program based on LLM

LLMarking: Adaptive Automatic Short-Answer Grading Using Large Language Models

Auto short answer grading using LLM

Miccai Feta2024 Challenge 3rd Award

Multi-stage segmentation solution

FDH-TTA: Fourier-Domain Harmonization for Test-Time Adaptation in Multicenter Medical Image Segmentation

Doman adaptation based on Fourier Transfer

Path Planning of Autonomous Pollination Using Heredity Algorithm for Intelligent Agriculture

Heredity Algorithm path planning for Pollination robot