Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Who Is Helping Whom? Analyzing Inter-Dependencies to Evaluate Cooperation in Human-AI Teaming
Upasana Biswas, Vardhan Palod, Siddhant Bhambri, Subbarao Kambhampati
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Pair ZSC agents with a scripted cooperative partner instead of relying on task reward alone to judge cooperation | COLE/MEP/HSP/FCP · Overcooked Counter Circuit (Non-RC) | Int_cons vs cooperative upper bound: COLE 0.60 vs 4.425; MEP 0.83 vs 6.59 | Hurt |
| Partner-initiated interdependencies: ZSC agents' acceptance | ZSC agents · Overcooked Counter Circuit (Non-RC), scripted partner | % triggered interdependencies not accepted: COLE 88.88, MEP 75.55, HSP 100.0, FCP 100.0 | Hurt |
| Compare reward and cooperation in required-cooperation (RC) vs non-required (Non-RC) layouts | ZSC agents with humans · Overcooked | correlation of task reward with Int_cons: RC r = 0.81; Non-RC r = 0.19 | Mixed |
| Add teaming reward (constructive interdependencies, alpha=0.3) to the training reward (ablation) | FCP, MEP, HSP · Overcooked | teaming reward during training | Improved |
| Add teaming reward to the training reward (ablation) | FCP, MEP, HSP · Overcooked | task reward | No clear effect |
| Higher task reward does not imply more cooperation in self-play Non-RC (COLE vs HSP) | COLE, HSP · Overcooked Counter Circuit (Non-RC), self-play | Int_cons at equal reward: both reward 120; Int_cons 10.23 vs 1.82 | Mixed |
Closest papers by meaning
- Partner Modelling Emerges in Recurrent Agents (But Only When It Matters) (NeurIPS 2025 · similarity 0.89)
- COOP$^2$: Defining, Observing, and Repairing Cooperation in LLM Multi-Agent Systems (NeurIPS 2026 · similarity 0.88)
- Designs for Enabling Collaboration in Human-Machine Teaming via Interactive and Explainable Systems (NeurIPS 2024 · similarity 0.88)
- Beyond Single Stationary Policies: Meta-Task Players as Naturally Superior Collaborators (NeurIPS 2024 · similarity 0.88)
- NestRL: A Nested Training Regime for Mutual Adaptation in Human–AI Teaming (NeurIPS 2026 · similarity 0.88)
- Adaptive Agents for Mixed-Initiative Human-AI Collaborations (AAAI 2025 · similarity 0.88)
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.
