Publications

61 papers, 40 of them peer-reviewed, with 290+ citations on Google Scholar. I led 24 of them as first, co-first or main hands-on author, and was senior author on 37, where I mentored the first author or led the design. Each title opens a page with the abstract, links and BibTeX.

All papers

Area
Status
My role
Year
61 papers

Safety Post-Training & Alignment

Mechanistic Interpretability & XAI

Agents & Evaluation

Tabular Foundation Models & Distillation

Climate Change & Earth Observation

Medical Imaging

Uncertainty & Robustness

NLP & AI for Social Good

Vision & Super-Resolution

Team releases

Libraries & tools

Designed & built v1, then mentored★ 144

TabTune

Unified inference, fine-tuning and benchmarking for tabular foundation models (21 models in 7 architecture families).

Led & built; design lead & mentor★ 49

AlignTune

Modular post-training toolkit for SFT, DPO, GRPO and RLHF with interchangeable backends; runs production alignment work in BFSI, legal and healthcare.

Designed & mentored★ 37

CuratorKIT

Provenance-grounded data curation and synthetic generation for LLM post-training.

Architected & built★ 28

DLBacktrace v2

torch.export-based, model-agnostic explainability for LLMs and MoEs with CUDA acceleration.

Led & built★ 32

CircuitKIT

Circuit discovery, evaluation and application toolkit for mechanistic interpretability.

Led & built★ 24

SafeTune

Audits and repairs safety drift in fine-tuned LLMs.

Built (first author)★ 18

xai_evals

Framework for evaluating post-hoc local explanation methods.

Team lead & contributor★ 19

AuditKIT

Source-available evaluation library for benchmarks, LLM-as-judge, RAG, model comparison, cost metrics and experiment tracking.

Designed, led & mentored★ 14

AgentTune

Source-available training lifecycle for building, training, evaluating and maintaining tool-using LLM agents.

Star counts update from GitHub on every site build.