Research map / Label-efficient and robust learning / Recognition and long-tail
Extracting Interaction-Aware Monosemantic Concepts in Recommender Systems
Dor Arviv, Yehonatan Elisha, Oren Barkan, Noam Koenigstein
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 prediction-level loss (increase beta from 0) (ablation) | SAE on MF and NCF (ML1M) · ML1M top-30 list fidelity | RBO and Kendall Tau | Improved |
| increase prediction-loss weight beta to large values (ablation) | SAE on MF and NCF (ML1M) · ML1M | monosemanticity score and sparsity | Hurt |
| add Matryoshka nested dictionaries (ablation) | SAE · ML1M | interpretability of neurons | No clear effect |
| add Matryoshka nested dictionaries (ablation) | SAE · Last.FM | neuron hierarchy (qualitative) | Improved |
| increase top-K from 10 to 50 | SAE concept neurons · ML1M and Last.FM purity | semantic purity: e.g. MF Country 1.00 -> 0.56 | Hurt |
| edit neuron activation in item latent vector (Bob Dylan) | MF on Last.FM · Last.FM top-30 recommendations | promotion to non-fan users | Improved |
Closest papers by meaning
- Actions Speak Louder than Words: Trillion-Parameter Sequential Transducers for Generative Recommendations (ICML 2024 · similarity 0.86)
- Semantic Convergence: Harmonizing Recommender Systems via Two-Stage Alignment and Behavioral Semantic Tokenization (AAAI 2025 · similarity 0.86)
- Language Representations Can be What Recommenders Need: Findings and Potentials (ICLR 2025 · similarity 0.86)
- A Geometric Approach to Personalized Recommendation with Set-Theoretic Constraints Using Box Embeddings (ICML 2025 · similarity 0.86)
- FACE: A General Framework for Mapping Collaborative Filtering Embeddings into LLM Tokens (NeurIPS 2025 · similarity 0.86)
- Improving LLMs for Recommendation with Out-Of-Vocabulary Tokens (ICML 2025 · similarity 0.86)
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.
