Research map / Reinforcement learning / Visual and representation RL
Complex Instruction Following with Diverse Style Policies in Football Games
Chenglu Sun, Shuo Shen, Haonan Hu, Wei Zhou, Chen Chen
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 ASaB to the Style Interpreter (ablation) | SI · Instruction-to-SP dataset | MAE: 0.752 -> 0.671 | Improved |
| add ASaB to the Style Interpreter (ablation) | SI · Instruction-to-SP dataset | inference time (s): 0.023 -> 0.094 | Hurt |
| use ASaB instead of matmul fusion (ablation) | SI · Instruction-to-SP dataset | MAE: 0.846 -> 0.671 | Improved |
| use ASaB instead of cross-attention (ablation) | SI · Instruction-to-SP dataset | MAE: 1.477 -> 0.671 | Improved |
| scale PLM from Qwen2.5-0.5B to Qwen2.5-3B (ablation) | SI · Instruction-to-SP dataset | MAE: 0.671 -> 0.629 | Improved |
| scale PLM from Qwen2.5-0.5B to Qwen2.5-3B (ablation) | SI · Instruction-to-SP dataset | inference time (s): 0.094 -> 0.785 | Hurt |
| use PSS instead of Uniform sampling | DST · GRF 5v5 style training | SEU: 1.81 -> 3.50 | Improved |
| use PSS instead of Uniform sampling | DST · GRF 5v5 style training | SMUL: 4.13 -> 2.44 | Improved |
| use PSS instead of Uniform sampling | DST · GRF 5v5 style training | SELO: 1029 -> 1142 | Improved |
| use PSS instead of LSDR sampling | DST · GRF 5v5 style training | SEU: 1.87 -> 3.50 | Improved |
| use PSS instead of LSDR sampling | DST · GRF 5v5 style training | SELO: 1048 -> 1142 | Improved |
| use SI instead of BC-Z | style translation · Instruction-to-SP dataset | MAE: 0.751 -> 0.671 | Improved |
Closest papers by meaning
- Policy Learning from Tutorial Books via Understanding, Rehearsing and Introspecting (NeurIPS 2024 · similarity 0.84)
- Prompted Policy Search: Reinforcement Learning through Linguistic and Numerical Reasoning in LLMs (NeurIPS 2025 · similarity 0.83)
- RLZero: Direct Policy Inference from Language Without In-Domain Supervision (NeurIPS 2025 · similarity 0.83)
- Vision-Based Generic Potential Function for Policy Alignment in Multi-Agent Reinforcement Learning (AAAI 2025 · similarity 0.83)
- The CoachAI Badminton Environment: A Novel Reinforcement Learning Environment with Realistic Opponents (Student Abstract) (AAAI 2024 · similarity 0.83)
- Rule Based Rewards for Language Model Safety (NeurIPS 2024 · similarity 0.82)
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.
