Research map / Reinforcement learning / Visual and representation RL
Multi-Objective Intrinsic Reward Learning for Conversational Recommender Systems
Zhendong Chu, Nan Wang, Hongning Wang
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 intrinsic-reward fine-tuning (CRSIRL) over policy-gradient pretraining (ablation) | CRSIRL · LastFM | success rate SR@15: 0.724 -> 0.772 | Improved |
| add intrinsic-reward fine-tuning (CRSIRL) over policy-gradient pretraining (ablation) | CRSIRL · Yelp* | success rate SR@15: 0.598 -> 0.622 | Improved |
| add intrinsic-reward fine-tuning (CRSIRL) over policy-gradient pretraining (ablation) | CRSIRL · LastFM* | hDCG@(15,10): 0.401 -> 0.431 | Improved |
| add hindsight reward shaping (HRS) (ablation) | CRSIRL · LastFM | average turns: 10.58 -> 10.12 | Improved |
| add hindsight reward shaping (HRS) (ablation) | CRSIRL · Yelp* | average turns: 11.21 -> 10.61 | Improved |
| add recommendation preference matching (RPM) (ablation) | CRSIRL · Yelp* | hDCG@(15,10): 0.213 -> 0.228 | Improved |
| add recommendation preference matching (RPM) (ablation) | CRSIRL · LastFM | success rate SR@15: 0.754 -> 0.772 | Improved |
| use multi-objective optimization instead of equal-weight multi-task learning (ablation) | CRSIRL · Yelp* | hDCG@(15,10): 0.207 -> 0.228 | Improved |
| use multi-objective optimization instead of equal-weight multi-task learning (ablation) | CRSIRL · LastFM | average turns: 10.09 vs 10.12 | No clear effect |
| equal-weight multi-task learning over policy gradient only (ablation) | PG · LastFM | success rate SR@15: 0.724 -> 0.768 | Improved |
| intrinsic reward coefficient lambda too small | CRSIRL · LastFM | performance | Hurt |
| intrinsic reward coefficient lambda too large | CRSIRL · LastFM | performance | Hurt |
Closest papers by meaning
- Enhancing Personalized Multi-Turn Dialogue with Curiosity Reward (NeurIPS 2025 · similarity 0.88)
- ConvRec-R1: Training LLM-based Conversational Recommender Systems with Reinforcement Learning (ICLR 2026 · similarity 0.87)
- Objective Shaping with Hard Negatives: Windowed Partial AUC Optimization for RL-based LLM Recommenders (NeurIPS 2026 · similarity 0.87)
- Think before Recommendation: Autonomous Reasoning-enhanced Recommender (NeurIPS 2025 · similarity 0.86)
- FacetCRS: Multi-Faceted Preference Learning for Pricking Filter Bubbles in Conversational Recommender System (AAAI 2024 · similarity 0.85)
- Capturing Individual Human Preferences with Reward Features (NeurIPS 2025 · similarity 0.85)
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.
