*indicating first authored publications

1. Characterizing Real-World Bugs in Tile Programs for Automated Bug Detection
   ACM SIGSOFT 35th International Symposium on Software Testing and Analysis (ISSTA 2026), 2026
   Ravishka Rathnasuriya*, Zihe Song*, Nidhi Majoju, Aaryaa Moharir, Tingxi Li, Wei Yang, Tao Xie

2. On-the-Fly Input Adaptation for Reliable Code Intelligence
   In The IEEE/ACM 48th International Conference on Software Engineering - New Idea and Emerging Result (ICSE-NIER 2026), 2026
   Ravishka Rathnasuriya*, Wei Yang

3. When to Answer and When to Defer: A Decision Framework for Reliable Code Predictions
   In The IEEE/ACM 48th International Conference on Software Engineering - New Idea and Emerging Result (ICSE-NIER 2026), 2026
   Ravishka Rathnasuriya*, Wei Yang

4. SoK: Efficiency Robustness of Dynamic Deep Learning Systems
   The 34th USENIX Security Symposium (USENIX Security 2025), 2025
   Ravishka Rathnasuriya*, Tingxi Li, Zexin Xu, Zihe Song, Mirazul Haque, Simin Chen, Wei Yang

5. An Investigation on Numerical Bugs in GPU Programs Towards Automated Bug Detection
   ACM SIGSOFT 34th International Symposium on Software Testing and Analysis (ISSTA 2025), 2025
   Ravishka Rathnasuriya*, Nidhi Majoju, Zihe Song, Wei Yang 

6. CodeImprove: Program Adaptation for Deep Code Models
   In The IEEE/ACM 47th International Conference on Software Engineering (ICSE 2025), 2025
   Ravishka Rathnasuriya*, Zijie Zhao, Wei Yang 

7. A Framework for On the Fly Input Refinement for Deep Learning Models
   Doctoral Symposium in the IEEE/ACM 47th International Conference on Software Engineering: Companion Proceedings (ICSE-Companion), 2025
   Ravishka Rathnasuriya*

8. On the Fly Input Refinement for Code Language Models
   Student Research Competition in the IEEE/ACM 47th International Conference on Software Engineering: Companion Proceedings (ICSE-Companion), 2025
   Ravishka Rathnasuriya*

9. Exploiting Efficiency Vulnerabilities in Dynamic Deep Learning Systems
   In The Proceedings of the 10th IEEE European Symposium on Security and Privacy-Posters (Euro S&P-Posters 2025), 2025
   Ravishka Rathnasuriya*, Wei Yang

10. Can you mimic me? Exploring the Use of Android Record & Replay Tools in Debugging
    In The 12th International Conference on Mobile Software Engineering and Systems (MOBILESoft 2025), 2025
    Zihe Song, S M Hasan Mansur, Ravishka Rathnasuriya, Yumna Fatima, Wei Yang, Kevin Moran, Wing Lam

11. COMET: Closed-loop Orchestration for Malicious Elicitation Techniques in Code Models
    In The 1st Amazon Nova AI Challenge Proceedings
    Zexin Xu, Tingxi Li, Ravishka Shemal Rathnasuriya, Zihe Song, Jun Ren, Bhavesh Mandalapu, Soroush Setayeshpour, Xinya Du, Wei Yang

12. HateModerate: Testing Hate Speech Detectors against Content Moderation Policies
    In the Annual Conference of the North American Chapter of the Association for Computational Linguistics -Findings. (NAACL 2024)
    Jiangrui Zheng, Xueqing Liu, Guanqun Yang, Mirazul Haque, Xing Qian, Ravishka Rathnasuriya, Girish Budhrani, Wei Yang
    
    

*indicating first authored submissions

1. Ravishka Rathnasuriya*, Wei Yang. Beyond Confidence: Rethinking Uncertainty Calibration in Deep Code Models 

2. Ravishka Rathnasuriya*, Wei Yang. Beyond Execution: Uncertainty Estimation and Abstention in LLM-Based Code Generation 

3. Tingxi Li, Mingfang Ji, Ravishka Rathnasuriya, Simin Chen, Yitao Hu, Wei Yang. Exploiting Efficiency Vulnerabilities in Vision-based Dynamic Deep Learning Pipeline Systems

*indicating first authored submissions

1. Ravishka Rathnasuriya*, Wei Yang. Semantic-Constrained Latent-Space Transformation for Input Adaptation in Code Models 

2. Ravishka Rathnasuriya*, Aarya Moharir, Nidhi Majoju, Wei Yang. Gradient-Guided Input Adaptation for Improving Image Classifier Performance 

3. Ravishka Rathnasuriya*, Aarya Moharir, Nidhi Majoju, Wei Yang. On-the-Fly Input Adaptation for Image Classifiers 

4. Ravishka Rathnasuriya*, Wei Yang. Improving Vision-Language Models via Input Adaptation 

5. Ravishka Rathnasuriya*, Wei Yang. TextImprove: Natural Language Adaptation for Language Models 

6. Ravishka Rathnasuriya*, Wei Yang. Evaluating the Efficiency Robustness of Large Language Models 

7. Ravishka Rathnasuriya*, Wei Yang. MOESlowDown: Understanding and Testing Efficiency Degradation of Mixture-of-Expert Models

8. Ravishka Rathnasuriya*, Wei Yang. Towards Foundation Models for Users with Customized Requirements: Bridging the Gap Between General Models and Specific Needs. 

9. Ravishka Rathnasuriya*, Simin Chen, Wei Yang. An Empirical Study on Automated Oracle Generation for Testing Deep Learning Application.

I am leading the following projects currently. 

1. Input Validation for DL/LLMs on Software Engineering-, Natural Language-, and Computer Vision- based tasks.

2. Input Refinement for DL/LLMs on Software Engineering-, Natural Language-, and Computer Vision- based tasks at deployment stage.

3. Selecting Effective Code Generations via Black Box Approach on Code LLMs.

4. Efficiency Robustness on Dynamic Deep Learning Systems (Mixture-of-Experts, Mixture-of-Depths, and Sparsity).

5. Program analysis on numerical bugs detection for GPU programs, differential/metamorphic testing for DL models.