Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Unified Prediction and Planning via Conflict-Aware Disjoint Parameter Training
Taewon Seo, Seonae Jeon, Giwon Lee, KUK-JIN YOON, Daehee Park
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 |
|---|---|---|---|
| add DPT to Task Arithmetic merging | Task Arithmetic · JRDB planning | ADE: 0.9387 -> 0.5865 | Improved |
| add DPT to Ties Merging | Ties Merging · JRDB planning | ADE: 0.8759 -> 0.4193 | Improved |
| add DPT to Sparse Merging | Sparse Merging · JRDB planning | ADE: 0.9246 -> 0.4044 | Improved |
| add DPT to Task Arithmetic | Task Arithmetic · JTA planning | ADE: 3.4726 -> 1.0373 | Improved |
| add DPT to Sparse Merging | Sparse Merging · JRDB prediction | ADE: 0.6587 -> 0.5952 | Improved |
| add DPT to plan fine-tuning | Plan Finetune · JRDB prediction | ADE: 1.9825 -> 0.7368 | Improved |
| add DPT to pred fine-tuning | Pred Finetune · JRDB prediction | ADE: 0.4967 -> 0.5250 | Hurt |
| add joint reasoning fine-tuning stage on inactive parameters | DPT + Sparse Merging · JRDB planning | ADE: 0.4044 -> 0.3640 | Improved |
| add joint reasoning stage | DPT + Sparse Merging · JRDB prediction | ADE: 0.5952 -> 0.5097 | Improved |
| raise mask ratio K from 2 to 40 in sparse merging (ablation) | DPT + Sparse Merging · JRDB planning | ADE: 0.4041 -> 0.4229 | Hurt |
| raise DPT allocation to planning from 50 to 90 percent (ablation) | DPT + Sparse Merging · JRDB planning | ADE: 0.3898 -> 0.4899 | Hurt |
| raise DPT allocation to prediction from 50 to 90 percent (ablation) | DPT + Sparse Merging · JRDB prediction | ADE: 0.5952 -> 0.6046 | Hurt |
Closest papers by meaning
- Interaction-Merged Motion Planning: Effectively Leveraging Diverse Motion Datasets for Robust Planning (ICCV 2025 · similarity 0.87)
- SOCIALGYM 2.0: Simulator for Multi-Robot Learning and Navigation in Shared Human Spaces (AAAI 2024 · similarity 0.86)
- Latent Variable Sequential Set Transformers for Joint Multi-Agent Motion Prediction (ICLR 2022 · similarity 0.85)
- SocialNav: Training Human-Inspired Foundation Model for Socially-Aware Embodied Navigation (CVPR 2026 · similarity 0.85)
- CE-Nav: Flow-Guided Reinforcement Refinement for Cross-Embodiment Local Navigation (ICLR 2026 · similarity 0.85)
- Social-Mamba: Socially-Aware Trajectory Forecasting with State-Space Models (ECCV 2026 · similarity 0.84)
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.
