paper-system/criteria.txt

44 lines
2.9 KiB
Plaintext

Criteria for Paper Selection:
When evaluating paper titles and abstracts, please consider the following criteria. Prioritize inclusiveness to reduce the risk of missing potentially interesting research.
Papers must meet ALL of these criteria:
1. Practical Applications: Does the paper focus on real-world or practical applications of Large Language Models (LLMs), particularly in areas like knowledge graphs, retrieval-augmented generation (RAG), or agentic AI?
2. Experimental Results and Quantitative Metrics: Does the paper include experimental results with quantitative metrics that demonstrate performance improvements or innovative applications? Look for concrete results related to areas like prompt engineering.
3. Comparison with State-of-the-Art: Does the paper compare its results or methodologies with existing state-of-the-art techniques, and does it demonstrate any advancements in terms of performance, scalability, or efficiency?
Papers should meet at least one of these criteria:
4. Methodology and Implementation Details: Does the paper clearly describe its methodology and implementation, particularly how LLMs are utilized in practical or real-world tasks?
5. Real-world Applications and Challenges: Does the paper discuss real-world applications or address limitations and challenges involving LLMs, especially in autonomous or agentic AI scenarios?
Additional Considerations:
- Novelty: Does the paper introduce novel approaches or techniques, especially those extending the application or integration of LLMs with technologies like knowledge graphs?
- Is the paper's approach implementable with current standard tools?
- Agentic AI: Does the paper describe applications where LLMs enable AI to autonomously perform complex, real-world tasks beyond sandboxed interactions?
- Reproducibility and Documentation: Are there sufficient details to support reproducibility and transparent documentation of the methodologies and results?
- Impact through Experimental Validation: Does the paper show robust experimental validation, with datasets and methods that closely reflect real-world scenarios?
The paper should be REJECTED if it:
- Primarily focuses on medical applications of AI
- Primarily focuses on social applications of AI in regard to Diversity, Social harm, or similar issues.
- Primarily focuses on video processing
- Primarily focuses on responsible AI application or AI ethics
- Primarily focuses on law, either with AI as subject or participant.
Instructions:
- Analyze each paper's title and abstract to determine how many criteria are met.
- Prioritize inclusiveness: Favor selection over rejection if a paper seems marginally relevant.
- Use a simple scoring system: A paper meeting at least two or three criteria should be accepted for further review.
Please make a best-guess decision based on the information provided in the titles and abstracts. Err on the side of potential inclusion to capture a broad spectrum of relevant research.