Scholar
Junlin Wang
Google Scholar ID: s1Qw2pAAAAAJ
Duke University
Citations & Impact
Citations
544
H-index
8
i10-index
6
Publications
15
Co-authors
9
list available
Contact
Publications
Resume (English only)
Academic Achievements
- Mixture-of-Agents Enhances Large Language Model Capabilities (arXiv, 2024)
- Reasoning in Token Economies: Budget-Aware Evaluation of LLM Reasoning Strategies (EMNLP, 2024)
- ReCaLL: Membership Inference via Relative Conditional Log-Likelihoods (EMNLP, 2024)
- Raccoon: Prompt Extraction Benchmark of LLM-Integrated Applications (ACL Findings, 2024)
- LLM-Resistant Math Word Problem Generation via Adversarial Attacks (EMNLP Findings, 2024)
- NeuroComparatives: Neuro-Symbolic Distillation of Comparative Knowledge (NAACL Findings, 2024)
- Maestro: A Gamified Platform for Teaching AI Robustness (EAAI, 2023)
Research Experience
- Worked closely with Sameer Singh on Machine Learning Interpretations and Natural Language Processing projects before joining Duke. Also worked as a research intern at several companies.
Education
- PhD in Computer Science at Duke University, advised by Bhuwan Dhingra; also advised by Sam Wiseman from 2022-2023.
Background
- Research Interests: LLM reasoning, agents, and alignment. Has interned at multiple companies including Together AI, AWS, Intel, Tencent, and the Applied AI lab at Comcast.
