Talks & events

Talks, posters and conferences. I give talks on post-training safety and interpretability; email me to invite me.

Upcoming

EMNLP 2026, Budapest, Hungary, October 24–29, 2026

Attending, with accepted work at the main conference, Findings, System Demonstrations and the NLLP Workshop. See the papers →

At conferences

2025.12EurIPS Spotlight Talk: Interpretability as Alignment: Making Internal Understanding a Design Principle, EurIPS Workshop on Private AI Governance, Copenhagen, with Aadit Sengupta
AryaXAI Alignment Lab webinar: Inside the Black Box: Interpreting LLMs with DL-Backtrace
AryaXAI Alignment Lab webinar: Beyond Explainability: Evaluating XAI Methods with Confidence Using xai_evals
Paper podcast: Interpretability-Aware Pruning in Medical Imagery

2026ICML 2026, Seoul, South Korea
2026The Web Conference (WWW) 2026, Dubai, UAE (attended remotely)
EurIPS Workshop on Private AI Governance, Copenhagen, Denmark (December 2025)
2025MICCAI 2025 Workshop, Daejeon, South Korea
2025ICML 2025, Vancouver, Canada
2025ICLR 2025, Singapore
2023AAAI 2023, Washington D.C., USA

2026Hindustan Times Brunch: shared perspectives with Vinay Kumar Sankarapu on AI sycophancy and why chatbots can sound confident even when they are wrong, in Taranna Khetpal's feature.
2026.02India AI Impact Summit 2026, Bharat Mandapam, New Delhi, representing Lexsi Labs. Attended AI Safety Connect Day (International Association for Safe and Ethical AI); joined The Dialogue's closed-door roundtable Powering the AI Frontier: Building India's Compute, Infrastructure, and Trust Backbone for Inclusive AI Growth; presented Lexsi Labs' work on alignment, interpretability-driven safety and open-source toolkits.
2026.02PyTorch Day Bangalore 2026
2025.12AurionAI Launch Event, India, representing Lexsi Labs
2024MLOps Conference, San Francisco, representing AryaXAI