Gradient attribution from differentiable AI weather models acts as a calibrated, near-optimal but gaming-vulnerable signal for allocating rewards based on data value in participatory weather sensing.
CrowdQC+—a quality-control for crowdsourced air-temperature ob- servations enabling world-wide urban climate applications
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Calibrating Attribution Proxies for Reward Allocation in Participatory Weather Sensing
Gradient attribution from differentiable AI weather models acts as a calibrated, near-optimal but gaming-vulnerable signal for allocating rewards based on data value in participatory weather sensing.