Research map / Architectures and efficiency / RNNs and normalization
Recommender Forest for Efficient Retrieval
Chao Feng, Wuchao Li, Defu Lian, Zheng Liu, Enhong Chen
This paper tried a common change: Add positional encoding
Add positional embedding (Abs) to decoder · Gowalla · NDCG@20: 0.1697 -> 0.2632 Improved
What 646 accepted papers report for this change
- 525 improved
- 58 no clear effect
- 7 mixed
- 56 hurt
- Learning-Augmented Scalable Linear Assignment Problem Optimization via Neural Dual Warm-Starts ICML 2026 · add positional encodings to features (d=13 -> d=21) · Block-structured, N=4096 · speedupImproved
- Mesh Based Simulations with Spatial and Temporal awareness ICML 2026 · use learned positional embedding, learnable relative bias or weighted adjacency instead of RoPE · Aneurysm · RMSEHurt
- Sparse ActionGen: Accelerating Diffusion Policy with Real-time Pruning ICML 2026 · add sinusoidal positional encoding to the pruner · Robomimic Tool Hang · success rate (%): 28 -> 54Improved
See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.
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 |
|---|---|---|---|
| Increase branch number from 2 to 4 (ablation) | RecForest single tree · Gowalla | NDCG@20: 0.2353 -> 0.2795 | Improved |
| Increase branch number from 2 to 18 (ablation) | RecForest single tree · Amazon | NDCG@20: 0.1624 -> 0.2088 | Improved |
| Increase branch number from 2 to 18 (ablation) | RecForest single tree · MIND | NDCG@20: 0.6900 -> 0.7154 | Improved |
| Increase branch number from 2 to 18 (ablation) | RecForest single tree · Yelp | NDCG@20: 0.1951 -> 0.2581 | Improved |
| Add positional embedding (Abs) to decoder (ablation) | RecForest · Gowalla | NDCG@20: 0.1697 -> 0.2632 | Improved |
| Add positional embedding (Abs) to decoder (ablation) | RecForest · MIND | NDCG@20: 0.6476 -> 0.6986 | Improved |
| Add positional embedding (Abs) to decoder (ablation) | RecForest · Amazon | NDCG@20: 0.1704 -> 0.1805 | Improved |
| Add positional embedding (Abs) to decoder (ablation) | RecForest · Yelp | NDCG@20: 0.1988 -> 0.2505 | Improved |
| Replace Abs positional embedding with RelK or RelKQ (ablation) | RecForest · Gowalla | NDCG@20: 0.2632 -> 0.2617 | No clear effect |
| Build forest with Random+Kmeans instead of Random or Kmeans (ablation) | RecForest forest · Amazon, MIND, Gowalla, Yelp | NDCG@20 | Improved |
| Build forest with Kmeans only (no randomness) (ablation) | RecForest forest · Amazon, MIND, Gowalla, Yelp | NDCG@20 | No clear effect |
| Build forest with purely Random trees (ablation) | RecForest forest · Amazon, MIND, Gowalla, Yelp | NDCG@20 | Hurt |
Closest papers by meaning
- Learning Deep Tree-Based Retriever for Efficient Recommendation: Theory and Method (TPAMI 2025 · similarity 0.89)
- SHARP-Distill: A 68× Faster Recommender System with Hypergraph Neural Networks and Language Models (ICML 2025 · similarity 0.85)
- ContextGNN: Beyond Two-Tower Recommendation Systems (ICLR 2025 · similarity 0.84)
- Generalization Error Bounds for Two-stage Recommender Systems with Tree Structure (NeurIPS 2024 · similarity 0.84)
- Can LLMs Outshine Conventional Recommenders? A Comparative Evaluation (NeurIPS 2025 · similarity 0.84)
- Improving LLMs for Recommendation with Out-Of-Vocabulary Tokens (ICML 2025 · 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.
