カレンダー
↑
サイトマップ
↑
wikiの書き方
↑
Search
AND
OR
↑
Menu
Tips
PC関係
論文投稿
セミナー
planet lunch
宇宙物理学セミナー
観測的宇宙論速報
初期宇宙・相対論速報
計算機
Mailing lists
UTAP web
VPN
はじめての人へ
ネットワークの設定
プリンタの設定
メール・メーリス
計算機係マニュアル
計算機資源
研究生活
AttendanceManager
事務連絡
MailingLists
Procedure for the new e-mail address
メーリングリスト
計算機係より
専攻メールへの移行について
昼食会議事録
↑
Latest Updates
計算機/ネットワークの設定/IP
2024-04-25 (木) 17:29:10
セミナー/宇宙物理学セミナー/2024-04-25
2024-04-24 (水) 15:12:09
計算機/プリンタの設定
2024-04-18 (木) 16:28:26
Menubarの編集
開始行:
[[UTAPwiki/セミナー/宇宙物理学セミナー]]
Speaker: Tilman Hartwig
Title: Be careful what you wish for: Reward Modelling in AI
Abstract:
A computer program will do what you say and not what you ...
everyday life of an astronomer, this can for example lead...
fitting results, which can be spotted and corrected with ...
intuition. However, this problem becomes more serious for...
intelligence with explicit, human-designed reward functio...
present examples from various scientific domains where an...
exploits the reward function, which leads to undesired be...
Finally, I will present Reward Modelling as a novel solut...
problem of specification learning.
終了行:
[[UTAPwiki/セミナー/宇宙物理学セミナー]]
Speaker: Tilman Hartwig
Title: Be careful what you wish for: Reward Modelling in AI
Abstract:
A computer program will do what you say and not what you ...
everyday life of an astronomer, this can for example lead...
fitting results, which can be spotted and corrected with ...
intuition. However, this problem becomes more serious for...
intelligence with explicit, human-designed reward functio...
present examples from various scientific domains where an...
exploits the reward function, which leads to undesired be...
Finally, I will present Reward Modelling as a novel solut...
problem of specification learning.
ページ名:
トップ
新規
一覧
単語検索
最終更新
ヘルプ
最終更新のRSS