Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Zero-Shot Assistance in Sequential Decision Problems
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 |
|---|---|---|---|
| Use advice while inferring the agent's reward and bias online, rather than leaving the agent unassisted | AIAD · Inventory management | cumulative discounted reward: 123.7 ± 11.7 -> 185.5 ± 8.5 | Helped |
| Use advice and online inference instead of IRL followed by automation | AIAD · Inventory management | cumulative discounted reward: 165.6 ± 9.0 -> 185.5 ± 8.5 | Helped |
| Add direct automation alongside advice | AIAD · Inventory management | cumulative discounted reward: 185.5 ± 8.5 -> 175.0 ± 8.3 | Hurt |
| Infer the agent's optimism/pessimism bias instead of assuming no bias (ablation) | AIAD · Inventory management | cumulative discounted reward: 175.5 ± 9.4 -> 185.5 ± 8.5 | Helped |
| Infer the agent's optimism/pessimism bias instead of assuming pessimism (ablation) | AIAD · Inventory management | cumulative discounted reward: 173.3 ± 9.2 -> 185.5 ± 8.5 | Helped |
| Infer the agent's optimism/pessimism bias instead of assuming optimism (ablation) | AIAD · Inventory management | cumulative discounted reward: 179.9 ± 8.8 -> 185.5 ± 8.5 | Helped |
| Infer anchoring bias online instead of assuming no agents have it (ablation) | AIAD · Day Trip Design | objective value | Helped |
| Infer anchoring bias online instead of assuming all agents have it (ablation) | AIAD · Day Trip Design | objective value | Helped |
| Add direct automation to advice to bypass actions blocked by anchoring bias | AIAD · Day Trip Design | objective value | Helped |
| Add direct automation alongside advice when agents have no anchoring bias | AIAD · Day Trip Design without biases | objective value | No clear effect |
Closest papers by meaning
- Learning to Make Adherence-aware Advice (ICLR 2024 · similarity 0.86)
- Joint Scoring Rules: Competition Between Agents Avoids Performative Prediction (AAAI 2025 · similarity 0.86)
- Emergent Risk Awareness in Rational Agents under Resource Constraints (NeurIPS 2025 · similarity 0.86)
- Agentick: A Unified Benchmark for General Sequential Decision-Making Agents (NeurIPS 2026 · similarity 0.85)
- Observation Interference in Partially Observable Assistance Games (ICML 2025 · similarity 0.85)
- Contracting with a Learning Agent (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.