Research map / Large language models / AI and society
Updates in Human-AI Teams: Understanding and Addressing the Performance/Compatibility Tradeoff
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| update AI to more accurate but incompatible error boundary | human-AI team (CAJA) · CAJA user study | team score | Hurt |
| compatible AI update (no new errors) | human-AI team (CAJA) · CAJA user study | team score | Helped |
| retrain on more data with plain log loss | logistic regression · Mortality | compatibility: 0.40 | Hurt |
| add new-error dissonance term to loss (increasing lambda_c) | LR and MLP · Recidivism, Credit Risk, Mortality | compatibility vs ROC | Mixed |
| imitation / strict-imitation dissonance instead of new-error dissonance (ablation) | LR and MLP · three domains | ROC at given compatibility | Hurt |
| increase number of human-visible features | human-AI team · CAJA user study | team score | Hurt |
| increase stochasticity of AI errors | human-AI team · CAJA user study | normalized score | Hurt |
Closest papers by meaning
- Position: Human Baselines in Model Evaluations Need Rigor and Transparency (With Recommendations & Reporting Checklist) (ICML 2025 · similarity 0.86)
- Position: Challenges and Future Directions of Data-Centric AI Alignment (ICML 2025 · similarity 0.86)
- Position: Towards Bidirectional Human-AI Alignment (NeurIPS 2025 · similarity 0.86)
- Automating Data Annotation under Strategic Human Agents: Risks and Potential Solutions (NeurIPS 2024 · similarity 0.86)
- Utilizing Human Behavior Modeling to Manipulate Explanations in AI-Assisted Decision Making: The Good, the Bad, and the Scary (NeurIPS 2024 · similarity 0.85)
- Human Expertise in Algorithmic Prediction (NeurIPS 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.