Research map / Large language models / Retrieval and knowledge
BEE-RAG: Balanced Entropy Engineering for Retrieval-Augmented Generation
Yuhao Wang, Ruiyang Ren, Yucheng Wang, Jing Liu, Xin Zhao, Hua Wu, Haifeng 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 balanced entropy factor β (ablation) | Zero-BEE · Zero-shot, avg EM (Table 2) | EM: 20.35 -> 31.42 | Improved |
| add mean µ of balancing factor (ablation) | Zero-BEE · Zero-shot, avg EM (Table 2) | EM: 27.47 -> 31.42 | Improved |
| add variance σ2 of balancing factor (ablation) | Zero-BEE · Zero-shot, avg EM (Table 2) | EM: 26.92 -> 31.42 | Improved |
| add parallel intrinsic multi-importance inference (ablation) | Zero-BEE · Zero-shot, avg EM (Table 2) | EM: 29.89 -> 31.42 | Improved |
| replace β with reranker scores (ablation) | Zero-BEE · Zero-shot, avg EM (Table 2) | EM: 31.42 -> 28.87 | Hurt |
| add residual in sentence-level vector (ablation) | Light-BEE · Fine-tuning, avg EM (Table 2) | EM: 23.65 -> 38.09 | Improved |
| add initialization constraint on β (ablation) | Light-BEE · Fine-tuning, avg EM (Table 2) | EM: 3.23 -> 38.09 | Improved |
| add FFN module (ablation) | Light-BEE · Fine-tuning, avg EM (Table 2) | EM: 38.09 -> 36.41 | Hurt |
| replace β with reranker scores (ablation) | Light-BEE · Fine-tuning, avg EM (Table 2) | EM: 38.09 -> 35.82 | Hurt |
| apply BEE zero-shot over vanilla RAG | Qwen-2.5-7B · Average of 4 QA datasets | EM: 28.65 -> 31.42 | Improved |
| apply Light-BEE over LoRA | Qwen-2.5-7B · Average of 4 QA datasets | EM: 33.70 -> 38.09 | Improved |
| prompt-based CoT / Chain-of-Note / Self-Critic over vanilla RAG | Qwen-2.5-7B · Average of 4 QA datasets | EM: 28.65 -> 28.89 (CoT) | No clear effect |
Closest papers by meaning
- RAGGED: Towards Informed Design of Scalable and Stable RAG Systems (ICML 2025 · similarity 0.90)
- Accelerating Inference of Retrieval-Augmented Generation via Sparse Context Selection (ICLR 2025 · similarity 0.90)
- Inference Scaling for Long-Context Retrieval Augmented Generation (ICLR 2025 · similarity 0.90)
- Long-Context LLMs Meet RAG: Overcoming Challenges for Long Inputs in RAG (ICLR 2025 · similarity 0.89)
- A Theory for Token-Level Harmonization in Retrieval-Augmented Generation (ICLR 2025 · similarity 0.89)
- Benchmarking Large Language Models in Retrieval-Augmented Generation (AAAI 2024 · similarity 0.89)
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.
