AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers propose Bellman-Taylor score decoding, a novel deep reinforcement learning framework designed to handle Markov decision processes with state-dependent action constraints common in operations research. The method decouples policy learning into a Euclidean score space while maintaining feasibility through an action decoder, enabling standard DRL algorithms to optimize complex systems like queueing networks without architectural modifications.
AIBullisharXiv – CS AI · Jun 106/10
🧠Researchers propose HMAF, a hierarchical allocation framework that optimizes ad impression distribution across guaranteed delivery contracts and real-time bidding auctions. Implemented at Meituan, the system achieved a 3.72% improvement in delivery rates and 1.59% revenue increase by unifying previously decoupled optimization approaches.
AINeutralarXiv – CS AI · Jun 45/10
🧠Researchers present an algorithmic framework for efficiently maintaining sheaf cohomology computations on dynamically evolving cellular complexes, reducing edit processing time from O(mn³) to O(1) per operation under bounded local geometry assumptions. The method demonstrates practical viability through experiments on large-scale graphs with millions of vertices and streaming edits, achieving microsecond-level latency while maintaining zero computational drift.
AIBullisharXiv – CS AI · May 116/10
🧠Researchers demonstrate that large language models can generate effective heuristics for hierarchical task network (HTN) planning, achieving near-optimal performance compared to state-of-the-art planners. LLM-generated heuristics reduce search effort on 83% of benchmark problems, suggesting AI models can enhance algorithmic planning efficiency beyond classical approaches.
AIBullisharXiv – CS AI · Apr 156/10
🧠Researchers propose an optimal model partitioning algorithm for split learning that reduces training delays by up to 38.95% by representing AI models as directed acyclic graphs and solving the problem via maximum-flow methods. The approach includes a low-complexity block-wise algorithm that achieves 13x faster computation on edge computing hardware, advancing the feasibility of distributed AI inference on mobile and edge devices.
🏢 Nvidia
AINeutralarXiv – CS AI · Mar 114/10
🧠Researchers developed a framework to identify what makes AI-generated optimal solutions more interpretable to humans, focusing on bin-packing problems. The study found that humans prefer solutions with three key properties: alignment with greedy heuristics, simple within-bin composition, and ordered visual representation.
AINeutralarXiv – CS AI · Mar 54/10
🧠Researchers propose Co-Evolutionary Alignment (CoEA), a new recommendation system method that uses dual large language models to balance relevant and novel content suggestions. The system addresses traditional recommendation bias through dynamic optimization that considers both long-term group identity and short-term individual preferences.
AINeutralarXiv – CS AI · Mar 44/103
🧠Researchers present a framework for social planners to strategically reveal positive and negative role models to influence agent behavior in social networks. The study addresses optimization challenges when disclosure budgets are limited and proposes algorithms to maximize social welfare while maintaining fairness across different groups.
