Papers & Artefacts
Every paper with its status — peer-reviewed, preprint or under submission — and links to the code and data needed to examine it.
Citation counts: Google Scholar, 149 in total as of 2026-10-07.
Register
Datasets, benchmarks and artefacts
Versioned code and data behind the papers, with licences and access conditions.
Method
Reproducibility packages
What should accompany a paper so another team can rerun it.
Free tool
Reproducibility checklist
Check whether your own study is ready for an independent rerun.
Peer-reviewed
A Four-Axis Trustworthiness Benchmark for LLM-as-Judge in Principle-Based Regulation
KDD 2026 Workshop on Secure and Trustworthy Large Language Models (SeT-LLM), poster
Dipankar Sarkar
From Errors to Proofs: Minimal-Core-Guided Repair for Neuro-Symbolic Constraint Solving
IJCAI-ECAI 2026 Workshop on Logic and Symbolic Reasoning (LogiSymb), poster
Dipankar Sarkar
Fed-Focal Loss for imbalanced data classification in Federated Learning
Workshop on Federated Learning for Data Privacy and Confidentiality in Conjunction with IJCAI 2020
Dipankar Sarkar , A Narang , S Rai
Preprints and manuscripts
Evaluating Bounded Autonomy in Regulated Agentic AI: A Diagnostic Harness with Constitutional Rewards, Escalation Labels, and Runtime Governance
arXiv preprint arXiv:2609.37501
Dipankar Sarkar
When Privacy Moves ML-Mediated Decisions On Device: Information and Incentive Misalignment in Auctions
arXiv preprint arXiv:2609.33312
Dipankar Sarkar
How Reproducible Are Evaluation Conclusions? A Self-Audit of LLM-Inferred Prompt Structure
arXiv preprint arXiv:2609.30074
Dipankar Sarkar
Operator-Aware Mixed-Precision Tolerance Calibration for Tensor Kernels
arXiv preprint arXiv:2607.16228
Dipankar Sarkar
Static PTX Metrics Track Structural Kernel Regressions but Miss Semantic Ones
arXiv preprint arXiv:2607.02541
Dipankar Sarkar
Test-Input Generation for Tensor Programs: What Actually Finds Kernel Bugs
arXiv preprint arXiv:2606.27396
Dipankar Sarkar
The Correctness Illusion in LLM-Generated GPU Kernels
arXiv preprint arXiv:2606.20128
Dipankar Sarkar
Before the Pull Request: Mining Multi-Agent Coordination
arXiv preprint arXiv:2606.19616
Dipankar Sarkar
An Input-Regime Audit of Conflict Detection for Retrieval-Augmented Generation
Manuscript submitted to the 2nd Workshop on Vector Databases (VecDB @ VLDB 2026)
Dipankar Sarkar
Epistral Network: Revolutionizing Media Curation and Consumption through Decentralization
arXiv preprint arXiv:2402.04881
Dipankar Sarkar , S Upadhyay
Navigating the Knowledge Sea: Planet-scale answer retrieval using LLMs
arXiv preprint arXiv:2402.05318
Dipankar Sarkar
FairFlow Protocol: Equitable Maximal Extractable Value (MEV) mitigation in Ethereum
arXiv preprint arXiv:2312.12654
Dipankar Sarkar
Viz: A QLoRA-based Copyright Marketplace for Legally Compliant Generative AI
arXiv preprint arXiv:2401.00503
Dipankar Sarkar
Centralized Intermediation in a Decentralized Web3 Economy: Value Accrual and Extraction
arXiv preprint arXiv:2311.08234
Dipankar Sarkar
Decentralized Deepfake Detection Blockchain Network using Dynamic Algorithm management
arXiv preprint arXiv:2311.18545
Dipankar Sarkar
Generalised DePIN Protocol: A Framework for Decentralized Physical Infrastructure Networks
arXiv preprint arXiv:2311.00551
Dipankar Sarkar
Towards Universal Atomic Composability: A Formal Model for Multi-Rollup Environments on Ethereum
arXiv preprint arXiv:2311.00422
Dipankar Sarkar
Curriculum generation using Autoencoder based continuous optimization
arXiv preprint arXiv:2106.08569
Dipankar Sarkar , M Gupta
One Shot Audio to Animated Video Generation
arXiv preprint arXiv:2102.09737
N Kumar , S Goel , A Narang , B Lall , M Hasan , P Agarwal , Dipankar Sarkar
Catfedavg: Optimising communication-efficiency and classification accuracy in federated learning
arXiv preprint arXiv:2011.07229
Dipankar Sarkar , S Rai , A Narang