Research map / Image generation / Fast diffusion sampling
Steering Diffusion Models Towards Credible Content Recommendation
Zhuo Cai, Shoujin Wang, Jin Li, Peilin Zhou, Victor W. Chu, Fang Chen, Tianqing Zhu, Charu C. Aggarwal
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 disentanglement module (ablation) | Disco · PolitiFact | HC@5: 0.3033 -> 0.3455 | Improved |
| add disentanglement module (ablation) | Disco · MHMisinfo | HC@5: 0.2650 -> 0.3000 | Improved |
| add credible subspace projection (ablation) | Disco · PolitiFact | CR@5: 0.9431 -> 0.9835 | Improved |
| add progressive enhancement of credible subspace (ablation) | Disco · PolitiFact | CR@5: 0.9423 -> 0.9835 | Improved |
| add preference contrast term (ablation) | Disco · GossipCop | HR@5: 0.4643 -> 0.5236 | Improved |
| use cosine error instead of MSE in the objective (ablation) | Disco · GossipCop | HC@5: 0.0010 -> 0.4918 | Improved |
| use item embedding rather than disentangled embedding as diffusion target (ablation) | Disco · GossipCop | HC@5: 0.4265 -> 0.4918 | Improved |
| use Disco instead of PreferDiff | PreferDiff · GossipCop | CR@5: 0.8307 -> 0.9277 | Improved |
| use Disco instead of DiffuRec | DiffuRec · PolitiFact | HC@5: 0.3334 -> 0.3466 | Improved |
| increase negative preference weight w continually (ablation) | Disco · PolitiFact, GossipCop, MHMisinfo | accuracy: best at w = 0.5, 1.5, 1 respectively | Hurt |
| reduce embedding dimension from 3072 to 64 (ablation) | DM-based recommenders · PolitiFact | accuracy | Hurt |
| use DDIM sampling without classifier-free guidance instead of DDPM | Disco · PolitiFact inference time | seconds: 994.3 -> 10.5 vs PRISM | Improved |
Closest papers by meaning
- Fading to Grow: Growing Preference Ratios via Preference Fading Discrete Diffusion for Recommendation (NeurIPS 2025 · similarity 0.87)
- On Efficiency-Effectiveness Trade-off of Diffusion-based Recommenders (NeurIPS 2025 · similarity 0.86)
- Learning Multiple User Distributions for Recommendation via Guided Conditional Diffusion (AAAI 2025 · similarity 0.85)
- Preference Diffusion for Recommendation (ICLR 2025 · similarity 0.85)
- Plug-In Diffusion Model for Sequential Recommendation (AAAI 2024 · similarity 0.84)
- CURE4Rec: A Benchmark for Recommendation Unlearning with Deeper Influence (NeurIPS 2024 · 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.
