Research map / Large language models / Language, safety and interpretability
What Generative Search Engines Like and How to Optimize Web Content Cooperatively
Yujiang Wu, Shanshan Zhong, Yubin Kim, Chenyan Xiong
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 |
|---|---|---|---|
| rewrite documents with AutoGEO_API rules instead of best manual baseline (Fluency Optimization) | Gemini generative engine · Researchy-GEO | GEO Overall: 27.75 -> 43.76 | Improved |
| train compact AutoGEO_Mini with RL on rule rewards instead of using the API model | Qwen3-1.7B rewriter · Researchy-GEO | GEO Overall: 27.75 (Fluency Opt.) -> 38.53 | Improved |
| add rule reward to AutoGEO_Mini RL (ablation) | AutoGEO_Mini · Researchy-GEO | GEO Overall: 31.41 -> 38.53 | Improved |
| add outcome reward to AutoGEO_Mini RL (ablation) | AutoGEO_Mini · Researchy-GEO | GEO Overall: 34.38 -> 38.53 | Improved |
| add semantic reward to AutoGEO_Mini RL (ablation) | AutoGEO_Mini · Researchy-GEO | GEO Overall: 37.79 -> 38.53 | Improved |
| include rule prompt in AutoGEO_Mini (ablation) | AutoGEO_Mini · Researchy-GEO | GEO Overall: 37.04 -> 38.53 | Improved |
| use AutoGEO instead of hijack/poisoning attacks to raise visibility | Gemini engine · Researchy-GEO | Insight (engine utility): 49.52 (Hijack) -> 53.79 | Improved |
| apply AutoGEO_API rewriting | Gemini engine · Researchy-GEO | Insight (engine utility): 51.07 -> 53.79 | Improved |
| apply AutoGEO_API on Claude engine (utility effect) | Claude engine · Researchy-GEO | Precision: 96.51 -> 84.98 | Hurt |
| apply AutoGEO_API on E-commerce Gemini engine | Gemini engine · E-commerce | Recall: 96.81 -> 94.46 | Hurt |
| transfer rule sets across engines or domains instead of engine-specific rules | AutoGEO_API · cross-engine/domain transfer | GEO score | Mixed |
| apply AutoGEO on low-visibility documents | Gemini engine · Researchy-GEO low-visibility | GEO Overall: 9.46 -> 35.83 | Improved |
Closest papers by meaning
- OneSearch-V2: The Latent Reasoning Enhanced Self-distillation Generative Search Framework (NeurIPS 2026 · similarity 0.85)
- EcoGEO: Trajectory-Aware Evidence Ecosystems for Web-Enabled LLM Search Agents (NeurIPS 2026 · similarity 0.85)
- Preference learning made easy: Everything should be understood through win rate (ICML 2025 · similarity 0.85)
- WEPO: Web Element Preference Optimization for LLM-based Web Navigation (AAAI 2025 · similarity 0.84)
- DynamicRAG: Leveraging Outputs of Large Language Model as Feedback for Dynamic Reranking in Retrieval-Augmented Generation (NeurIPS 2025 · similarity 0.84)
- Learning to Rank in Generative Retrieval (AAAI 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.
