New approach for AI models deciding when to ask experts for help

Original: Density-Ratio Losses for Post-Hoc Learning to Defer

Source: arxiv.org ↗