Can your database system predict underprovisining before it even happens?
Meet ◒ xBound, the very first framework for join size lower bounds. xBound tells you how many tuples your SQL query will produce *at least*.
Brought to you by @microsoft.com Gray Systems Lab & @utndatasystems.bsky.social.
Meet ◒ xBound, the very first framework for join size lower bounds. xBound tells you how many tuples your SQL query will produce *at least*.
Brought to you by @microsoft.com Gray Systems Lab & @utndatasystems.bsky.social.
January 21, 2026 at 9:40 AM
Can your database system predict underprovisining before it even happens?
Meet ◒ xBound, the very first framework for join size lower bounds. xBound tells you how many tuples your SQL query will produce *at least*.
Brought to you by @microsoft.com Gray Systems Lab & @utndatasystems.bsky.social.
Meet ◒ xBound, the very first framework for join size lower bounds. xBound tells you how many tuples your SQL query will produce *at least*.
Brought to you by @microsoft.com Gray Systems Lab & @utndatasystems.bsky.social.
This, and also:
Randomly spaced announcements declaring "single car train service is now in effect, please board..." doesn't work. People stand at the 2nd car area and rush.
Do it like they do in the UK.
"The Xbound train to Y is approaching. This train is made of one coach. Please board..."
Randomly spaced announcements declaring "single car train service is now in effect, please board..." doesn't work. People stand at the 2nd car area and rush.
Do it like they do in the UK.
"The Xbound train to Y is approaching. This train is made of one coach. Please board..."
Ottawa needs marked boarding zones for the single Lints, as well as an audio and visual announcement. I understand using single Lints to save a few bucks, but in it's current state it's an accessibility issue.
April 20, 2025 at 5:56 PM
This, and also:
Randomly spaced announcements declaring "single car train service is now in effect, please board..." doesn't work. People stand at the 2nd car area and rush.
Do it like they do in the UK.
"The Xbound train to Y is approaching. This train is made of one coach. Please board..."
Randomly spaced announcements declaring "single car train service is now in effect, please board..." doesn't work. People stand at the 2nd car area and rush.
Do it like they do in the UK.
"The Xbound train to Y is approaching. This train is made of one coach. Please board..."
Mihail Stoian, Tiemo Bang, Hangdong Zhao, Jes\'us Camacho-Rodr\'iguez, Yuanyuan Tian, Andreas Kipf: xBound: Join Size Lower Bounds https://arxiv.org/abs/2601.13117 https://arxiv.org/pdf/2601.13117 https://arxiv.org/html/2601.13117
January 21, 2026 at 6:30 AM
Mihail Stoian, Tiemo Bang, Hangdong Zhao, Jes\'us Camacho-Rodr\'iguez, Yuanyuan Tian, Andreas Kipf: xBound: Join Size Lower Bounds https://arxiv.org/abs/2601.13117 https://arxiv.org/pdf/2601.13117 https://arxiv.org/html/2601.13117
1/ Paper: arxiv.org/abs/2601.13117
xBound: Join Size Lower Bounds
Cloud database vendors invest substantial resources into their query optimizers, and for good reason. Cardinality estimation, a cornerstone of the optimizer, is critical for the selection of efficient...
arxiv.org
January 21, 2026 at 9:40 AM
1/ Paper: arxiv.org/abs/2601.13117
XBOUND evaluates device‑control agents per UI state; the 7B UI‑TARS model achieved the highest state‑level accuracy while models under 7B lag behind in tests. Read more: https://getnews.me/xbound-introduces-state-level-evaluation-for-device-control-agents/ #xbound #devicecontrol #uitars
September 29, 2025 at 10:03 PM
XBOUND evaluates device‑control agents per UI state; the 7B UI‑TARS model achieved the highest state‑level accuracy while models under 7B lag behind in tests. Read more: https://getnews.me/xbound-introduces-state-level-evaluation-for-device-control-agents/ #xbound #devicecontrol #uitars
of DC agents in mastering these states. Furthermore, we have developed a ``pseudo'' episode tree dataset derived from Android Control test data. Utilizing this dataset and XBOUND, we comprehensively evaluate the OS-Atlas and UI-TARS series, examining [5/7 of https://arxiv.org/abs/2505.21279v1]
May 28, 2025 at 5:59 AM
of DC agents in mastering these states. Furthermore, we have developed a ``pseudo'' episode tree dataset derived from Android Control test data. Utilizing this dataset and XBOUND, we comprehensively evaluate the OS-Atlas and UI-TARS series, examining [5/7 of https://arxiv.org/abs/2505.21279v1]
proposing the XBOUND evaluation method, which employs the calculation of a novel Explore Metric to delineate the capability boundaries of DC agents. Compared to previous evaluation methods, XBOUND focuses on individual states to assess the proficiency [4/7 of https://arxiv.org/abs/2505.21279v1]
May 28, 2025 at 5:59 AM
proposing the XBOUND evaluation method, which employs the calculation of a novel Explore Metric to delineate the capability boundaries of DC agents. Compared to previous evaluation methods, XBOUND focuses on individual states to assess the proficiency [4/7 of https://arxiv.org/abs/2505.21279v1]
Zhang, Chen, Zhang, Li, Weng, Xiang, Nie, Zhang: XBOUND: Exploring the Capability Boundaries of Device-Control Agents through Trajectory Tree Exploration https://arxiv.org/abs/2505.21279 https://arxiv.org/pdf/2505.21279 https://arxiv.org/html/2505.21279
May 28, 2025 at 5:59 AM
Zhang, Chen, Zhang, Li, Weng, Xiang, Nie, Zhang: XBOUND: Exploring the Capability Boundaries of Device-Control Agents through Trajectory Tree Exploration https://arxiv.org/abs/2505.21279 https://arxiv.org/pdf/2505.21279 https://arxiv.org/html/2505.21279
