Research map / Reinforcement learning / Visual and representation RL
UNO! UNified Offline Training Paradigm for Learning Path Recommendation
Linzhi Peng, Wentao Zhu, Ke Cheng, Heng Chang, Junchen Ye, Bowen Du, Weifeng Lv
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 |
|---|---|---|---|
| train UniLPR with GRPO and dense process reward instead of sparse-reward PPO (ablation) | UNO / UniLPR · Junyi his-20 | Ep (%): 17.35 -> 34.32 | Improved |
| train UniLPR with GRPO and dense process reward instead of sparse-reward AC (ablation) | UNO / UniLPR · Assist09 his-20 | Ep (%): 13.81 -> 35.08 | Improved |
| use UniLPR backbone instead of SASRec with GRPO (ablation) | UNO / UniLPR · Junyi his-20 | Ep (%): 26.83 -> 34.32 | Improved |
| use UniLPR backbone instead of GRU4Rec with GRPO (ablation) | UNO / UniLPR · Assist09 exp-20 | Ep (%): 2.16 -> 8.66 | Improved |
| add process reward and GRPO to SASRec on Assist09 exploratory (ablation) | UNO / UniLPR · Assist09 exp-20 | Ep (%): -2.68 -> 8.24 | Improved |
| UNO vs baselines across rct/his/exp tasks (ablation) | UNO / UniLPR · Junyi, Assist09 | Ep (%): Junyi rct-10 46.01 vs 37.64 (PPO) | Improved |
| remove KS loss, KL loss or personalized advantage (ablation) | UNO / UniLPR · rct-10, his-20, exp-20 | Ep (%) | Improved |
Closest papers by meaning
- Personalized Exercise Recommendation with Semantically-Grounded Knowledge Tracing (NeurIPS 2025 · similarity 0.85)
- GenAL: Generative Agent for Adaptive Learning (AAAI 2025 · similarity 0.84)
- Objective Shaping with Hard Negatives: Windowed Partial AUC Optimization for RL-based LLM Recommenders (NeurIPS 2026 · similarity 0.83)
- Multi-type MOOCs Recommendation: Leveraging Deep Multi-Relational Representation and Hierarchical Reasoning (AAAI 2025 · similarity 0.83)
- CURE4Rec: A Benchmark for Recommendation Unlearning with Deeper Influence (NeurIPS 2024 · similarity 0.83)
- Would You Like Your Data to Be Trained? A User Controllable Recommendation Framework (AAAI 2024 · similarity 0.83)
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.
