-{"basics":{"name":"Keyu He","label":"Master in NLP@CMU SCS || NLP Researcher","image":"keyu.jpg","email":"keyuhe@cmu.edu","phone":"(213) 713-2973","url":"https://keyu-he.github.io/","summary":"Master of Intelligent Information Systems (MIIS) student at Carnegie Mellon University. Previously earned B.S. from USC with double major in Computer Science and Applied & Computational Mathematics. Passionate about socially intelligent AI systems and human-centered AI evaluation.","location":{"city":"Pittsburgh","region":"Pennsylvania","countryCode":"US"},"profiles":[]},"education":[{"institution":"Carnegie Mellon University","location":"Pittsburgh, PA","url":"https://www.cmu.edu/","area":"Master of Intelligent Information Systems (MIIS)","studyType":"Master's","startDate":"2025-08-01","endDate":"2027-05-01","score":"","summary":"Graduate program focusing on intelligent information systems, NLP research, and AI systems. Conducting research on explainable AI and socially intelligible AI.","courses":["Advanced Natural Language Processing (Grade TBD)","Introduction to Question Answering (Grade TBD)","Independent Study in Language Technologies (Grade TBD)","Directed Study in Language Technologies (Grade TBD)"]},{"institution":"University of Southern California","location":"Los Angeles, CA","url":"https://www.usc.edu/","area":"Computer Science and Applied & Computational Mathematics","studyType":"Bachelor's","startDate":"2021-08-01","endDate":"2025-05-01","score":"GPA: 3.98/4.00","summary":"Minor in Artificial Intelligence Applications. USC Dornsife Dean's List & USC Viterbi Dean's List (2021-2025). USC Center for Undergraduate Research in Viterbi Engineering (CURVE) fellow with $6,750 in research funding (Spring 2024 - Spring 2025). USC Academic Achievement Award covering $24,000 in tuition.","courses":["Language Models in Natural Language Processing (A)","Applied Machine Learning for Natural Language Processing (A)","Applications of Machine Learning (A)","Applied Neural Networks (A)","Introduction to Artificial Intelligence (A)","Capstone: Design and Construction of Large Software Systems (A)","Software Engineering (A)","Mathematical Statistics (A)","Probability Theory (A)","Linear Algebra and Differential Equations (A)","Numerical Methods (A)","Data Structure and Object-Oriented Design (A)","Introduction to Algorithms and the Theory of Computing (A-)"]}],"publications":[{"name":"Believing without Seeing: Quality Scores for Contextualizing Vision-Language Model Explanations","author":"Keyu He, Tejas Srinivasan, Brihi Joshi, Xiang Ren, Jesse Thomason, Swabha Swayamdipta","publisher":"Submitted to NeurIPS 2025","releaseDate":"2025-05","summary":"We introduce Visual Fidelity and Contrastiveness -- two explanation quality scores that help users more appropriately rely on vision-language model predictions without seeing the image."},{"name":"ELI-Why: Evaluating the Pedagogical Utility of LLM Explanations","author":"Brihi Joshi*, <b>Keyu He*</b>, Sahana Ramnath, Sadra Sabouri, Kaitlyn Zhou, Souti Chattopadhyay, Swabha Swayamdipta, Xiang Ren","publisher":"Findings of ACL 2025","releaseDate":"2025-02","summary":"Evaluate the pedagogical utility of LLMs in tailoring explanations to users with different educational backgrounds.","url":"../assets/pdf/ELI_Why_Evaluating_the_Pe.pdf"},{"name":"Attributing Culture-Conditioned Generations to Pretraining Corpora","author":"Huihan Li*, Arnav Goel*, Keyu He, Xiang Ren","publisher":"ICLR 2025","releaseDate":"2025-01","summary":"This paper introduces MEMOed, a framework to analyze whether AI generations are driven by memorization or generalization, with a focus on cultural symbols.","url":"../assets/pdf/Attributing_Culture_Cond.pdf"},{"name":"Enhancing Debugging Skills of LLMs with Prompt Engineering","author":"Keyu He*, Max Li*, and <b>Joseph Liu*</b>","publisher":"Technical Report","releaseDate":"2023-12","url":"../assets/pdf/Enhancing-Debugging.pdf","summary":"Study on improving LLM debugging through prompt engineering techniques. Evaluated few-shot learning and chain-of-thought approaches on GPT-3.5, revealing limitations in current debugging capabilities."}],"projects":[{"name":"LLM Prompt Recovery","summary":"Developed a system to recover user prompts using fine-tuned Mixtral models, achieving top 3.4% in a Kaggle competition.","startDate":"2024-03-01","endDate":"2024-04-01","highlights":["Achieved a score of 0.65 using sentence-T5-base and sharpened cosine similarity.","Ranked 75/2175 globally in the Kaggle competition.","Published fine-tuned model on Kaggle for broader access."],"url":"https://www.kaggle.com/models/keyuhe02/mixtral-8x7b-instruct-finetuned"},{"name":"AI-Based Career Advisor","summary":"Built an AI tool to assist users in planning career paths based on skills and interests.","startDate":"2024-11-01","endDate":"2024-12-01","highlights":["Developed a Streamlit-based interactive UI integrating GPT-4o for job suggestions.","Implemented cosine similarity search for skill-job matching.","Integrated Bing AI for real-time job application link retrieval."]},{"name":"Enhancing Debugging Skills of LLMs with Prompt Engineering","summary":"Improved LLM debugging capabilities through advanced prompt engineering techniques.","startDate":"2023-08-01","endDate":"2023-11-01","highlights":["Experimented with various prompting strategies to enhance debugging efficiency.","Achieved significant improvements in LLM performance on debugging tasks."]},{"name":"Automated Hate Speech Detection in Social Media","summary":"Developed an advanced ML model for detecting hate speech, achieving 94% accuracy.","startDate":"2023-09-01","endDate":"2023-12-01","highlights":["Fine-tuned BERT for classification tasks.","Enhanced online safety and inclusivity through robust model optimization."]}],"skills":[{"name":"Programming","level":"","keywords":["C++","Python","Java","MySQL","HTML","CSS","JS","x86-64 Assembly"]},{"name":"Software","level":"","keywords":["PyTorch","Pandas","NumPy","Git","AWS","LaTeX","Mathematica","Matlab"]},{"name":"Areas of Expertise","level":"","keywords":["Machine Learning","Natural Language Processing (NLP)","Large Language Models (LLMs)","Data Science / Data Engineering"]},{"name":"Languages","level":"","keywords":["Mandarin (native)","English (professional)"]}],"awards":[{"title":"Silver Medal, Kaggle Competition","date":"2024-04","summary":"Ranked 75/2175 (Top 3.4%) on the global leaderboard, LLM Prompt Recovery Project.","url":"https://www.kaggle.com/models/keyuhe02/mixtral-8x7b-instruct-finetuned"},{"title":"USC Academic Achievement Award","date":"2022-12","summary":"Awarded in Fall 2022, Spring 2023, Spring 2024, and Fall 2024. Covered 11 units of tuition costs in total, amounting to approximately $24,000."},{"title":"4th Place, USC Integral Bee Competition","date":"2022-04","summary":"Ranked 4th in the USC Integral Bee Competition."},{"title":"1st Prize, International Linguistics Olympiad","date":"2021-07","summary":"Senior Level, Individual Open Round, China."},{"title":"1st Prize, International Linguistics Olympiad","date":"2021-07","summary":"Senior Level, Team Open Round, China."}]}
0 commit comments