Research map / Learning theory and algorithms / Prediction and decision losses
Approximating Drift-Diffusion Models for User Decisions under Nudging and External Information
Gustavo Grivol, Hanna Halaburda, Alexander Tuzhilin
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 |
|---|---|---|---|
| use ExtDDM (2-step with EGMN) instead of best baseline | ExtDDM + EGMN · KuaiRec 2.0 | KL divergence: 0.3601 -> 0.2063 | Improved |
| use ExtDDM + EGMN instead of CREAD | ExtDDM + EGMN · KuaiRec 2.0 | MAE: 4.3863 -> 4.2833 | Improved |
| use ExtDDM + EGMN instead of CREAD | ExtDDM + EGMN · KuaiRec 2.0 | XAUC: 0.5883 -> 0.6053 | Improved |
| use hybrid CREAD+EGMN with the same 66% quantile split (ablation) | CREAD+EGMN · KuaiRec 2.0 | MAE: 4.4597 (EGMN) -> 4.4432 | No clear effect |
| use ExtDDM instead of CREAD | ExtDDM · MOOCCubeX | MAE: 35.202 -> 32.342 | Improved |
| use ExtDDM instead of EGMN | ExtDDM · MOOCCubeX | KL divergence: 2.4245 -> 1.3453 | Improved |
| use ExtDDM instead of Intensity-Free at high threshold | ExtDDM · Simulated reflected DDM, u=12 | log-likelihood: -10.95 -> -9.80 | Improved |
| use ExtDDM instead of Intensity-Free at low threshold | ExtDDM · Simulated single-boundary DDM, u=4 | log-likelihood: -7.49 (IntFree) vs -7.53 (ExtDDM) | No clear effect |
| use ExtDDM instead of RMTPP/NHP | ExtDDM · Simulated DDMs | log-likelihood: -530.45 (RMTPP) -> -9.80 (reflected, u=12) | Improved |
| remove dynamic drift from ExtDDM (fixed threshold only) (ablation) | ExtDDM · KuaiRec 2.0 | performance gains | Hurt |
Closest papers by meaning
- How Distributed Collaboration Influences the Diffusion Model Training? A Theoretical Perspective (ICML 2025 · similarity 0.82)
- SyncSDE: A Probabilistic Framework for Diffusion Synchronization (CVPR 2025 · similarity 0.82)
- Learning from Streaming Data when Users Choose (ICML 2024 · similarity 0.82)
- Diffusion Models Meet Contextual Bandits (NeurIPS 2025 · similarity 0.82)
- CleanDiffuser: An Easy-to-use Modularized Library for Diffusion Models in Decision Making (NeurIPS 2024 · similarity 0.82)
- Identifying Predictions That Influence the Future: Detecting Performative Concept Drift in Data Streams (AAAI 2025 · 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.
