Offloading Score, a new measure of AI over-reliance, earns NeurIPS 2026 Oral
Diyi_Yang · x · 2026-09-25
Researchers introduce Offloading Score, a metric defining AI reliance as the fraction of cognitive effort offloaded to AI while completing a task. In a controlled user study, it detects increased reliance under time pressure where several common alternative measures fail.
The work tackles the growing worry that AI tools make users overreliant by making it measurable, and has been accepted as an Oral at the expanded NeurIPS 2026 Evaluations & Datasets track.
More from Research
- NVIDIA open-sources MONAI Physio: cardiac and respiratory digital twins from medical images — Paimaamu · 2026-09-25
- Stochastic Attention accepted at NeurIPS: samples attention at inference for calibrated uncertainty — Fowe · 2026-09-25
- Lila Sciences' AI finds palladium catalyst a 12-year expert rejected, running 1,000+ hours — AndrewLBeam · 2026-09-25
- Stanford professor on AI-run bio experiments: without motion control, novices can't follow the AI — suragnair · 2026-09-25
- New paper shows models can learn general patterns from zero real-world data — mark_k · 2026-09-25
- Mathematician's 20-Year Lessons: Edge-Case ML Failures Are Math Problems — TivadarDanka · 2026-09-25