Publications
12 papers since 2024, at NeurIPS, AAAI and ICML workshops and on arXiv. * marks shared first authorship. Each title opens the paper’s page, with the abstract, links and BibTeX. Also on Google Scholar.
2026
- Lost in Serialization: Invariance and Generalization of LLM Graph Reasoners
D. Herbst*, L. Karbevska*, D. Kumar*, A. Ahuja*, F. G. Nasrabadi*, F. Frasca · GCLR @ AAAI 2026; GFM @ ICML 2026co-first - Black-Box Red Teaming of Agentic AI: A Taxonomy-Driven Framework for Automated Risk Discovery
D. Kumar, N. A. Birur, T. Baswa, S. Agarwal, P. Harshangi · LAMaS Workshop @ AAAI 2026first author - SocioEval: A Template-Based Framework for Evaluating Socioeconomic Status Bias in Foundation Models
D. Kumar, I. Gupta, N. A. Birur, T. Baswa, S. Agarwal, P. Harshangi · 3rd Workshop on AI Governance (AIGOV) at AAAI 2026first author
2025
- Beyond Text: Multimodal Jailbreaking of Vision-Language and Audio Models through Perceptually Simple Transformations
D. Kumar*, S. Jena*, N. A. Birur, T. Baswa, S. Agarwal, P. Harshangi · NeurIPS Reliable ML from Unreliable Data Workshopco-first - Quantifying CBRN Risk in Frontier Models
D. Kumar, N. A. Birur, T. Baswa, S. Agarwal, P. Harshangi · NeurIPS Reliable ML from Unreliable Data Workshopfirst author - Beyond Western Politics: Cross-Cultural Benchmarks for Evaluating Partisan Associations in LLMs
D. Kumar*, I. Gupta*, N. A. Birur, T. Baswa, S. Agarwal, P. Harshangi · NeurIPS LLM Evaluation Workshopco-first - No Free Lunch with Guardrails
D. Kumar, N. A. Birur, T. Baswa, S. Agarwal, P. Harshangi · arXiv preprintfirst author
2024
- Investigating Implicit Bias in Large Language Models: A Large-Scale Study of Over 50 LLMs
D. Kumar*, U. Jain*, S. Agarwal, P. Harshangi · NeurIPS Safe Generative AI Workshopco-first - SAGE-RT: Synthetic Alignment data Generation for Safety Evaluation and Red Teaming
A. Kumar*, D. Kumar*, J. Loya, N. A. Birur, T. Baswa, S. Agarwal, P. Harshangi · NeurIPS Red Teaming GenAI Workshopco-first - Efficacy of the SAGE-RT Dataset for Model Safety Alignment: A Comparative Study
T. Baswa, N. A. Birur, D. Kumar, J. Loya, A. Kumar, P. Harshangi, S. Agarwal · NeurIPS Pluralistic Alignment Workshop - VERA: Validation and Enhancement for Retrieval Augmented Systems
N. A. Birur, T. Baswa, D. Kumar, J. Loya, S. Agarwal, P. Harshangi · arXiv preprint - Increased LLM Vulnerabilities from Fine-tuning and Quantization
D. Kumar, A. Kumar, S. Agarwal, P. Harshangi · arXiv preprintfirst author