Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..

Comparator-Adaptive $\Phi$-Regret: Improved Bounds, Simpler Algorithms, and Applications to Games

Authors: Soumita Hait, Ping Li, Haipeng Luo, Mengxiao Zhang

NeurIPS 2025 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical This paper is theoretic-focused and does not include experiments.
Researcher Affiliation Academia Soumita Hait University of Southern California EMAIL Ping Li Shanghai University of Finance and Economics EMAIL Haipeng Luo University of Southern California EMAIL Mengxiao Zhang University of Iowa EMAIL
Pseudocode Yes Algorithm 1 MWU over Φb with prior π Input: learning rate η > 0 and prior distribution π defined in Definition 3.2. Initialize q1 as π. for t = 1, 2 . . . , T do Propose ϕt = Eϕ qt[ϕ] S and receive loss matrix ptℓ t [0, 1]d d. Update qt+1 such that qt+1(ϕ) qt(ϕ) exp η ϕ, ptℓ t .
Open Source Code No This paper is theoretic-focused and does not include experiments.
Open Datasets No This paper is theoretic-focused and does not include experiments.
Dataset Splits No This paper is theoretic-focused and does not include experiments.
Hardware Specification No This paper is theoretic-focused and does not include experiments.
Software Dependencies No This paper is theoretic-focused and does not include experiments.
Experiment Setup No This paper is theoretic-focused and does not include experiments.