Matches in SemOpenAlex for { <https://semopenalex.org/work/W2901707424> ?p ?o ?g. }
- W2901707424 abstract "One obstacle to applying reinforcement learning algorithms to real-world problems is the lack of suitable reward functions. Designing such reward functions is difficult in part because the user only has an implicit understanding of the task objective. This gives rise to the agent alignment problem: how do we create agents that behave in accordance with the user's intentions? We outline a high-level research direction to solve the agent alignment problem centered around reward modeling: learning a reward function from interaction with the user and optimizing the learned reward function with reinforcement learning. We discuss the key challenges we expect to face when scaling reward modeling to complex and general domains, concrete approaches to mitigate these challenges, and ways to establish trust in the resulting agents." @default.
- W2901707424 created "2018-11-29" @default.
- W2901707424 creator A5003314770 @default.
- W2901707424 creator A5008987732 @default.
- W2901707424 creator A5020224050 @default.
- W2901707424 creator A5025651820 @default.
- W2901707424 creator A5029025914 @default.
- W2901707424 creator A5090592321 @default.
- W2901707424 date "2018-11-19" @default.
- W2901707424 modified "2023-09-27" @default.
- W2901707424 title "Scalable agent alignment via reward modeling: a research direction." @default.
- W2901707424 cites W145787216 @default.
- W2901707424 cites W1527197079 @default.
- W2901707424 cites W1541356759 @default.
- W2901707424 cites W1581742186 @default.
- W2901707424 cites W1583697287 @default.
- W2901707424 cites W1624823515 @default.
- W2901707424 cites W1777239053 @default.
- W2901707424 cites W1845972764 @default.
- W2901707424 cites W1855989615 @default.
- W2901707424 cites W1919575688 @default.
- W2901707424 cites W1921523184 @default.
- W2901707424 cites W1986014385 @default.
- W2901707424 cites W1999874108 @default.
- W2901707424 cites W2008645545 @default.
- W2901707424 cites W2061562262 @default.
- W2901707424 cites W2098441518 @default.
- W2901707424 cites W2101105183 @default.
- W2901707424 cites W2108677974 @default.
- W2901707424 cites W2113909127 @default.
- W2901707424 cites W2114864106 @default.
- W2901707424 cites W2121110499 @default.
- W2901707424 cites W2121863487 @default.
- W2901707424 cites W2137938746 @default.
- W2901707424 cites W2154023516 @default.
- W2901707424 cites W2154579312 @default.
- W2901707424 cites W2155277832 @default.
- W2901707424 cites W2156869222 @default.
- W2901707424 cites W2160371091 @default.
- W2901707424 cites W2162651021 @default.
- W2901707424 cites W2173248099 @default.
- W2901707424 cites W2205847592 @default.
- W2901707424 cites W2290104316 @default.
- W2901707424 cites W2323086467 @default.
- W2901707424 cites W2341475188 @default.
- W2901707424 cites W2432004435 @default.
- W2901707424 cites W2434014514 @default.
- W2901707424 cites W2462906003 @default.
- W2901707424 cites W2523568623 @default.
- W2901707424 cites W2531327146 @default.
- W2901707424 cites W2558801266 @default.
- W2901707424 cites W2574075983 @default.
- W2901707424 cites W2575120333 @default.
- W2901707424 cites W2580300496 @default.
- W2901707424 cites W2594475271 @default.
- W2901707424 cites W2594877703 @default.
- W2901707424 cites W2604117713 @default.
- W2901707424 cites W2616964725 @default.
- W2901707424 cites W2617191410 @default.
- W2901707424 cites W2618097077 @default.
- W2901707424 cites W2619787335 @default.
- W2901707424 cites W2620059021 @default.
- W2901707424 cites W2621205314 @default.
- W2901707424 cites W2626804490 @default.
- W2901707424 cites W2626967530 @default.
- W2901707424 cites W2640329709 @default.
- W2901707424 cites W2736601468 @default.
- W2901707424 cites W2749747771 @default.
- W2901707424 cites W2750866482 @default.
- W2901707424 cites W2760057500 @default.
- W2901707424 cites W2768908787 @default.
- W2901707424 cites W2771657877 @default.
- W2901707424 cites W2781474777 @default.
- W2901707424 cites W2782980316 @default.
- W2901707424 cites W2785324569 @default.
- W2901707424 cites W2786036274 @default.
- W2901707424 cites W2786104118 @default.
- W2901707424 cites W2786928559 @default.
- W2901707424 cites W2787708942 @default.
- W2901707424 cites W2790698955 @default.
- W2901707424 cites W2792012198 @default.
- W2901707424 cites W2792641098 @default.
- W2901707424 cites W2796885425 @default.
- W2901707424 cites W2797527950 @default.
- W2901707424 cites W2798877128 @default.
- W2901707424 cites W2799058376 @default.
- W2901707424 cites W2803392236 @default.
- W2901707424 cites W2803974723 @default.
- W2901707424 cites W2805516822 @default.
- W2901707424 cites W2808155616 @default.
- W2901707424 cites W2810469995 @default.
- W2901707424 cites W2810602713 @default.
- W2901707424 cites W2885183727 @default.
- W2901707424 cites W2890954269 @default.
- W2901707424 cites W2892375480 @default.
- W2901707424 cites W2895560838 @default.
- W2901707424 cites W2896930824 @default.
- W2901707424 cites W2903075639 @default.
- W2901707424 cites W2906586541 @default.
- W2901707424 cites W2908064123 @default.