Papers by Kiran Purohit
From Tokens to Steps: Verification-Aware Speculative Decoding for Efficient Multi-Step Reasoning (2026.findings-acl)
Copied to clipboard
| Challenge: | Speculative decoding (SD) allows a lightweight draft model to propose outputs that a stronger target model verifies. |
| Approach: | They propose a verification-aware speculative decoding framework that performs step-level verification using only model-internal signals. |
| Outcome: | Experiments show that SpecGuard outperforms both SD and reward-guided SD in accuracy and reliability tests. |
EXPLORA: Efficient Exemplar Subset Selection for Complex Reasoning (2024.emnlp-main)
Copied to clipboard
Kiran Purohit, Venktesh V, Raghuram Devalla, Krishna Yerragorla, Sourangshu Bhattacharya, Avishek Anand
| Challenge: | Recent advances in large language models (LLMs) have enabled in-context learning (ICL) a critical challenge in ICL is the selection of optimal exemplars . |
| Approach: | They propose an algorithm for static exemplar subset selection for reasoning tasks . they propose a method that estimates parameters without incorporating confidence information . |
| Outcome: | The proposed method significantly reduces the number of LLM calls to 11% of those required by state-of-the-art methods and achieves a substantial performance improvement of 12.24%. |