DD
Debojyoti Dutta
cs.LGcs.AIcs.SEcs.CLcs.NIstat.MLComputational Biologycs.CYcs.DCcs.IR
On Valency
published · living versionsW_b5vy6b9q·v1 · currentpublished
CR-Bench-v2: Evaluating the Real-World Utility of AI Code Review Agents
with Kristen Pereira, Neelabh Sinha, Rajat Ghosh
1 version
Preprints & journals
33 papers in the corpus · 2006–2026GRPO Training Dynamics for Small Language Models2609.39321v1 · Rajat Ghosh, Vaishnavi Bhargava, Henry Wong et al.2026 · 0 citationsarXiv
Predictable GRPO: A Closed-Form Model of Training Dynamics2606.30789v2 · Rajat Ghosh, Datta Nimmaturi, Aryan Singhal et al.2026 · 0 citationsarXiv
CR-Bench: Evaluating the Real-World Utility of AI Code Review Agents2603.11078v1 · Kristen Pereira, Neelabh Sinha, Rajat Ghosh et al.2026 · 0 citationsarXiv
SWE-Tester: Training Open-Source LLMs for Issue Reproduction in Real-World Repositories2601.13713v2 · Aditya Bharat Soni, Rajat Ghosh, Vaishnavi Bhargava et al.2026 · 0 citationsarXiv
Action Shapley: A Training Data Selection Metric for World Model in Reinforcement Learning2601.10905v1 · Rajat Ghosh, Debojyoti Dutta2026 · 0 citationsarXiv
Go-UT-Bench: A Fine-Tuning Dataset for LLM-Based Unit Test Generation in Go2511.10868v2 · Yashshi Pipalani, Hritik Raj, Rajat Ghosh et al.2025 · 0 citationsarXiv
A Multi-Agent Framework for Stateful Inference-Time Search2510.07147v1 · Arshika Lalan, Rajat Ghosh, Aditya Kolsur et al.2025 · 0 citationsarXiv
RANGER -- Repository-Level Agent for Graph-Enhanced Retrieval2509.25257v1 · Pratik Shah, Rajat Ghosh, Aryan Singhal et al.2025 · 0 citationsarXiv
BAR Conjecture: the Feasibility of Inference Budget-Constrained LLM Services with Authenticity and Reasoning2507.23170v2 · Jinan Zhou, Rajat Ghosh, Vaishnavi Bhargava et al.2025 · 0 citationsarXiv
Predictive Scaling Laws for Efficient GRPO Training of Large Reasoning Models2507.18014v3 · Datta Nimmaturi, Vaishnavi Bhargava, Rajat Ghosh et al.2025 · 0 citationsarXiv
MLKV: Efficiently Scaling up Large Embedding Model Training with Disk-based Key-Value Storage2504.01506v1 · Yongjun He, Roger Waleffe, Zhichao Han et al.2025 · 1 citationarXiv
AILuminate: Introducing v1.0 of the AI Risk and Reliability Benchmark from MLCommons2503.05731v2 · Shaona Ghosh, Heather Frase, Adina Williams et al.2025 · 4 citationsarXiv
CPP-UT-Bench: Can LLMs Write Complex Unit Tests in C++?2412.02735v1 · Vaishnavi Bhargava, Rajat Ghosh, Debojyoti Dutta2024 · 2 citationsarXiv
Efficient Alignment of Large Language Models via Data Sampling2411.10545v2 · Amrit Khera, Rajat Ghosh, Debojyoti Dutta2024 · 0 citationsarXiv
Introducing v0.5 of the AI Safety Benchmark from MLCommons2404.12241v2 · Bertie Vidgen, Adarsh Agrawal, Ahmed M. Ahmed et al.2024 · 5 citationsarXiv
DMLR: Data-centric Machine Learning Research -- Past, Present and Future2311.13028v2 · Luis Oala, Manil Maskey, Lilith Bat-Leah et al.2023 · 7 citationsarXiv
Data-Driven Evaluation of Training Action Space for Reinforcement Learning2204.03840v1 · Rajat Ghosh, Debojyoti Dutta2022 · 0 citationsarXiv
MLPerf Training Benchmark1910.01500v3 · Peter Mattson, Christine Cheng, Cody Coleman et al.2019 · 206 citationsarXiv
Foreign English Accent Adjustment by Learning Phonetic Patterns1807.03625v1 · Fedor Kitashov, Elizaveta Svitanko, Debojyoti Dutta2018 · 9 citationsarXiv
Detecting fraudulent activity in a cloud using privacy-friendly data aggregates1411.6721v1 · Marc Solanas, Julio Hernandez-Castro, Debojyoti Dutta2014 · 7 citationsarXiv
Optimal bandwidth-aware VM allocation for Infrastructure-as-a-Service1202.3683v1 · Debojyoti Dutta, Michael Kapralov, Ian Post et al.2012 · 5 citationsarXiv
Widescope - A social platform for serious conversations on the Web1111.1958v1 · Noah Burbank, Debojyoti Dutta, Ashish Goel et al.2011 · 2 citationsarXiv
Similarity Search and Locality Sensitive Hashing using TCAMs1006.3514v1 · Rajendra Shinde, Ashish Goel, Pankaj Gupta et al.2010 · 4 citationsarXiv
CSI: A Paradigm for Behavior-oriented Delivery Services in Mobile Human Networks0807.1153v1 · Wei-jen Hsu, Debojyoti Dutta, Ahmed Helmy2008 · 18 citationsarXiv
Counting clusters using R-NN curves.17602604 · Guha, Rajarshi, Dutta, Debojyoti, Wild, David J et al.2007 · 2 citationsJournal of chemical information and modeling. 2007;47(4):1308-18
MSNovo: a dynamic programming algorithm for de novo peptide sequencing via tandem mass spectrometry.17550227 · Mo, Lijuan, Dutta, Debojyoti, Wan, Yunhu et al.2007 · 96 citationsAnalytical chemistry. 2007;79(13):4870-8
Ensemble feature selection: consistent descriptor subsets for multiple QSAR models.17407280 · Dutta, Debojyoti, Guha, Rajarshi, Wild, David et al.2007 · 50 citationsJournal of chemical information and modeling. 2007;47(3):989-97
Speeding up tandem mass spectrometry database search: metric embeddings and fast near neighbor search.17237061 · Dutta, Debojyoti, Chen, Ting2007 · 48 citationsBioinformatics (Oxford, England). 2007;23(5):612-8
Local similarity analysis reveals unique associations among marine bacterioplankton species and environmental factors.16882654 · Ruan, Quansong, Dutta, Debojyoti, Schwalbach, Michael S et al.2006 · 295 citationsBioinformatics (Oxford, England). 2006;22(20):2532-8
R-NN curves: an intuitive approach to outlier detection using a distance based method.16859303 · Guha, Rajarshi, Dutta, Debojyoti, Jurs, Peter C et al.2006 · 19 citationsJournal of chemical information and modeling. 2006;46(4):1713-22
Local lazy regression: making use of the neighborhood to improve QSAR predictions.16859315 · Guha, Rajarshi, Dutta, Debojyoti, Jurs, Peter C et al.2006 · 62 citationsJournal of chemical information and modeling. 2006;46(4):1836-47
Mining Behavioral Groups in Large Wireless LANscs/0606002v2 · Wei-jen Hsu, Debojyoti Dutta, Ahmed Helmy2006 · 59 citationsarXiv
Scalable partitioning and exploration of chemical spaces using geometric hashing.16426067 · Dutta, Debojyoti, Guha, Rajarshi, Jurs, Peter C et al.2006 · 20 citationsJournal of chemical information and modeling. 2006;46(1):321-33
Career total: 88 works. 33 are in this corpus.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.