Shashank Kirtania

Bio

Hi! I am a Research Fellow (Pre-Doctoral) at Microsoft Research India. Previously, I was a Research Fellow with the program synthesis group (PROSE) at Microsoft in Bengaluru. I work on AI for Code and AI for Math. My research goal is to make automated programming more trustworthy by developing AI systems that generate robust, correct, and verifiable code—leveraging techniques from Formal Methods, Software Engineering, and Artificial Intelligence to ensure the reliability and correctness of automatically generated programs.

I have been privileged to work with Aditya Kanade, Ashish Tiwari, Gustavo Soares, Arjun Radhakrishna, Sumit Gulwani, and Arun Iyer.

I am applying for PhD. positions for fall 2026, in space of AI4Code, AI4Math and trustworthy ML. I will be at NeurIPS 2025, do reach out if you want to chat!!

For fun, I enjoy bouldering, running and playing chess.

Publications

Most recent publications on Google Scholar.
^‡ indicates equal contribution.

Selected
All

Activation Steering in Theorem Prover LLMs

Shashank Kirtania, Arun Iyer

FM in Wild Workshop ICLR'25, Under Submission at ICLR 2026

Paper

STACKFEED: Structured Textual Actor-Critic Knowledge Base Editing with Feedback

Shashank Kirtania‡, Naman Gupta‡, Priyanshu Gupta, Krishna Kariya, Sumit Gulwani, Arun Iyer, Suresh Parthasarathy, Arjun Radhakrishna, Sriram K. Rajamani, Gustavo Soares

Empirical Methods in Natural Language Processing 2025, DL for Code Workshop NeurIPS 2025

Paper

Improving Language Agents with BREW:

Shashank Kirtania, Param Biyani, Priyanshu Gupta, Yasharth Bajpai, Roshni Iyer, Sumit Gulwani, Gustavo Soares

Multi-Turn Interaction Workshop at NeurIPS 2025, Under Submission at ICLR 2026

Paper

MetaReflection: Learning Instructions for Language Agents using Past Reflections

Shashank Kirtania‡, Priyanshu Gupta‡, Annanya Singha‡, Sumit Gulwani, Arjun Radhakrishna, Sherry Shi, Gustavo Soares

Empirical Methods in Natural Language Processing 2024

Paper

LOGIC-LM++: Multi-Step Refinement for Symbolic Formulations

Shashank Kirtania, Priyanshu Gupta, Arjun Radhakrishna

Workshop on Natural Language Reasoning at ACL 2024

Paper

Activation Steering in Theorem Prover LLMs

Shashank Kirtania, Arun Iyer

FM in Wild Workshop ICLR'25, Under Submission at ICLR 2026

Paper

STACKFEED: Structured Textual Actor-Critic Knowledge Base Editing with Feedback

Shashank Kirtania‡, Naman Gupta‡, Priyanshu Gupta, Krishna Kariya, Sumit Gulwani, Arun Iyer, Suresh Parthasarathy, Arjun Radhakrishna, Sriram K. Rajamani, Gustavo Soares

Empirical Methods in Natural Language Processing 2025, DL for Code Workshop NeurIPS 2025

Paper

Improving Language Agents with BREW:

Shashank Kirtania, Param Biyani, Priyanshu Gupta, Yasharth Bajpai, Roshni Iyer, Sumit Gulwani, Gustavo Soares

Multi-Turn Interaction Workshop at NeurIPS 2025, Under Submission at ICLR 2026

Paper

MetaReflection: Learning Instructions for Language Agents using Past Reflections

Shashank Kirtania‡, Priyanshu Gupta‡, Annanya Singha‡, Sumit Gulwani, Arjun Radhakrishna, Sherry Shi, Gustavo Soares

Empirical Methods in Natural Language Processing 2024

Paper

LOGIC-LM++: Multi-Step Refinement for Symbolic Formulations

Shashank Kirtania, Priyanshu Gupta, Arjun Radhakrishna

Workshop on Natural Language Reasoning at ACL 2024

Paper

DWT-CompCNN: Deep Image Classification Network for High Throughput JPEG 2000 Compressed Documents

Tejasvee Bisen, Mohammed Javed, Shashank Kirtania, P Nagabhushan

Pattern Analysis and Applications, Springer, 2023.

Paper

Shashank Kirtania

Bio

Publications

CV