Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Boundedly rational agent modeling
Models human behavior with bias, bounded inference, beliefs, and epistemic planning to assist agents.
87 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Human-AI and goal-conditioned, keyword cluster “communication · ai · collaboration”
State History Tried Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
28 papers in 2023–24 and 24 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 9.5% to 6.2%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2016 · 'Knowing Whether' in Proper Epistemic Knowledge Bases (AAAI 2016 · 14 citations)
- 2017 · Repeated Inverse Reinforcement Learning (NeurIPS 2017 · 28 citations)
- 2018 · Strategic Coalitions With Perfect Recall (AAAI 2018)
- 2019 · Verification of RNN-Based Neural Agent-Environment Systems (AAAI 2019 · 46 citations)
- 2020 · Learning and Reasoning for Robot Sequential Decision Making under Uncertainty (AAAI 2020 · 25 citations)
- 2021 · Content Masked Loss: Human-Like Brush Stroke Planning in a Reinforcement Learning Painting Agent (AAAI 2021 · 29 citations)
- 2022 · Teaching Humans When to Defer to a Classifier via Exemplars (AAAI 2022 · 26 citations)
- 2023 · Do the Rewards Justify the Means? Measuring Trade-Offs Between Rewards and Ethical Behavior in the Machiavelli Benchmark (ICML 2023 · 26 citations)
- 2024 · Goal Alignment: Re-analyzing Value Alignment Problems Using Human-Aware AI (AAAI 2024 · 12 citations)
- 2025 · A Continuous-time Tractable Model for Present-biased Agents (AAAI 2025 · 1 citation)
- 2026 · Mini Amusement Parks (MAPs): A Testbed for Modelling Business Decisions (NeurIPS 2026 · 0 citations)
Earliest papers here
- 'Knowing Whether' in Proper Epistemic Knowledge Bases (AAAI 2016 · 14 citations)
- Modeling Human Understanding of Complex Intentional Action with a Bayesian Nonparametric Subgoal Model (AAAI 2016 · 8 citations)
- Model Checking Probabilistic Knowledge: A PSPACE Case (AAAI 2016 · 3 citations)
What has been tried
61 papers here report 447 measured changes (a component added, removed or swapped, compared with a baseline): helped 67%, no clear effect 9%, hurt 18%, mixed 6%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| learn the recommendation order instead of using a fixed preference-correlation order | Chengdu taxi simulator · Norm. TDI | 109.89% -> 109.97% | Helped | i-Rebalance: Personalized Vehicle Repositioning for Supply Demand Balance (AAAI 2024 · 7 citations) |
| Reinitialize a model's weight when a newly entering model replaces an older model | Taxi · adaptation speed | Helped | Dynamic Knowledge Injection for AIXI Agents (AAAI 2024 · 1 citation) | |
| Add language conditioning to the vision-only policy | Inference trajectories · action prediction accuracy | 0.86 -> 0.92 | Helped | MARPLE: A Benchmark for Long-Horizon Inference (NeurIPS 2024 · 1 citation) |
| Remove the assumed relationship between proxy and observed utility (alpha=0) | Medical AI, Z=1 intervention · preference-gap lower bound | -0.64 and -0.82 -> -1 | Hurt | The Limits of Predicting Agents from Behaviour (ICML 2025 · 0 citations) |
| Increase the coefficient of the simulator-loss term in the prover's loss | Graph isomorphism · message-sequence distribution difference | Helped | Neural Interactive Proofs (ICLR 2025 · 0 citations) | |
| Add multi-agent debate rather than proposing evidence in isolation | Sepsis preference recovery with confuser, L=6 · judge accuracy | 38\% to about 85\% | Helped | Reward Design for Justifiable Sequential Decision-Making (ICLR 2024 · 0 citations) |
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 52 papers; from OpenAlex and the papers' affiliations).
- Google DeepMind · 4 papers
- Microsoft · 3 papers
- Harvard University · 2 papers
- York University · 2 papers
- NTT Corporation · 2 papers
- CNRS · 2 papers
- National Institute of Advanced Industrial Science and Technology · 2 papers
- University of Oxford · 2 papers
- Nanjing University of Science and Technology · 2 papers
- Southeast University · 2 papers
Use it
Download BibTeX of 39 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Human-AI coordination agents (125 papers)
- Collaborative multi-agent perception (86 papers)
- Theory-of-mind social reasoning (69 papers)
- Emergent communication in referential games (50 papers)
Elsewhere in the topic
- World-model-based agents (110 papers)
- Imitation and skill learning (100 papers)
- Language-grounded reinforcement agents (96 papers)
- Embodied agent learning (93 papers)
- Intrinsically motivated exploration (86 papers)
- Interactive multi-agent trajectory prediction (75 papers)
- Learning-guided MAPF search (73 papers)
- Vision-language navigation agents (66 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).