I recently graduated as the Institute Silver Medalist (Rank 2) from BITS Pilani, where I majored in Computer Science and Engineering. I work at the intersection of multimodal machine learning, language models, interpretability, and more recently, agentic AI systems. I recently worked on GameDevBench (ICML 2026), where we evaluated agentic game development capabilities of frontier models.

For my undergraduate thesis at CMU with Prof. Graham Neubig, Simran Khanuja, and Michael Saxon, I built a system that augments vision-language models with multilingual, multicultural knowledge to better assess cultural relevance in images; this work was accepted to CEGIS @ ICCV 2025 and EACL 2026 (Main). Earlier, I explored ML applications in mental health through speech analysis and developed post-training methods for language models on nuanced, human-centered tasks.

I am currently exploring full-time opportunities starting in December 2026.

1. GameDevBench: Evaluating Agentic Capabilities Through Game Development

   Wayne Chi, Yixiong Fang, Arnav Yayavaram, Siddharth Yayavaram, Seth Karten, Qiuhong Anna Wei, Runkun Chen, Alexander Wang, Valerie Chen, Ameet Talwalkar, Chris Donahue

   Accepted at ICML 2026

2. CAIRE: Cultural Attribution of Images by Retrieval-Augmented Evaluation

   Arnav Yayavaram, Siddharth Yayavaram, Simran Khanuja, Michael Saxon, and Graham Neubig

   Published in CEGIS @ ICCV, 2025 and EACL, 2026

   Paper / Code

3. BERT-based Idiom Identification using Language Translation and Word Cohesion

   Arnav Yayavaram, Siddharth Yayavaram, Prajna Upadhyay, and Apurba Das

   Published in MWE-UD @ LREC-COLING, 2024

   Paper / Code

4. On the Effects of Concordance and Diversity in Example Selection for Adaptive In-Context Learning

   Manogna Nadella, Suryansh Chandola, Arnav Yayavaram, Siddharth Yayavaram, Kishan Abijay P, Siddharth Vijay, Varun Ravichandran, Priten Rathore, Prajna Upadhyay

   Under review at ACM SIGIR 2026

5. Interpretable Feature Optimization for Sadness Recognition in Speech Emotion Analysis

   Siddharth Yayavaram, Arnav Yayavaram, Jabez Christopher, and Vasan Arunachalam

   Published in IEEE Intelligent Systems, 2024

   Paper / Code