Featured3 projects
msmarco-genqa
active

Reproducible retrieve → rerank → generate pipeline on MS MARCO with paired-bootstrap evaluation and a manifest-enforced reproducibility contract.

retrieval · rag · evaluation · reproducibility·updated 2026-07-28
View project
rag-observatory
active

Research prototype for trace-based RAG observability and failure analysis.

rag · observability · failure-analysis · reproducibility·updated 2026-07-27
View project
Q-Learning Exploitability
maintained

Exact agent evaluation and failure analysis for self-play Q-learning.

reinforcement-learning · agent-evaluation · failure-analysis · reproducibility·updated 2026-09-05
View project
Related research1 project