Related Experiment Video
Updated: Aug 6, 2026
03:14
Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
Published on: December 6, 2024
Toward advanced management of sludge bulking: An explainable evaluation framework to quantify and improve large
Boyan Xu1, Ning Fan1, Chuankun Zhang1
1School of Technology for Sustainability, Beijing Normal University, China.
Water Research
|July 23, 2026
Summary
This study introduces an explainable framework to evaluate large language models (LLMs) for wastewater treatment. Integrating vision and retrieval-augmented generation (RAG) significantly improved LLM performance in managing activated sludge bulking.
Area of Science:
- Wastewater Treatment Engineering
- Artificial Intelligence in Environmental Science
- Microbial Ecology in Water Systems
Background:
- Activated sludge bulking is a common operational failure in wastewater treatment plants (WWTPs), requiring expert intervention often limited by specialist scarcity.
- Large language models (LLMs) offer potential for supporting bulking management, but lack a framework for evaluating their cognitive strengths and weaknesses.
Purpose of the Study:
- To develop an explainable evaluation framework to quantify LLMs' capabilities in understanding, analysis, and practice for sludge bulking management.
- To identify LLM limitations and guide improvements for advanced wastewater treatment decision-support.
Main Methods:
- Established an open-access evaluation suite with 280 tasks and an LLM-rater with ground truth for reliable assessment.
- Diagnosed LLMs' cognitive capabilities across understanding, analysis, and practice, evaluating performance based on task type and LLM identity.
- Integrated a vision module and retrieval-augmented generation (RAG) to enhance LLM performance, particularly for visual recognition and practical application.
Main Results:
- LLMs demonstrated limited understanding of microscopy-based recognition (<54% accuracy) and microbial identification (28-42% normalized scores).
- Analysis of process-microbe interactions showed weak performance (36-52% normalized scores), with practice tasks being particularly challenging.
- Integration of a vision module improved filament recognition (51% to 78% accuracy), and RAG enhanced DeepSeek-R1's practice capability by 69%.
Conclusions:
- The developed framework provides a scientific foundation for evaluating LLMs in sludge bulking management.
- Post-training pathways like RAG and vision modules can selectively enhance LLM cognitive capabilities, democratizing expertise for WWTPs.
- The RAG-enhanced DeepSeek-R1 (FilamentGPT) demonstrated practical feasibility for WWTP decision-support, achieving high overall performance.
Related Concept Videos
01:27Language and Cognition
Language serves as a bridge between ideas and communication, influencing how individuals perceive and interact with the world. Psychologists have long debated whether language shapes thought or vice versa. This discussion gained grip with Edward Sapir and Benjamin Lee Whorf in the 1940s, who proposed that language determines thought, a concept known as linguistic determinism. They suggested that the vocabulary and structure of a language influence how its speakers think and perceive reality.
01:28Automatic Processing and Automatic Social Behavior
Automatic processing refers to the cognitive operations that occur without conscious intent or awareness, playing a fundamental role in shaping social cognition and behavior. These processes enable individuals to navigate complex social environments efficiently by relying on mental shortcuts and pre-existing knowledge structures known as schemas. One of the most influential mechanisms underlying automatic processing is priming, which subtly activates mental representations through exposure to...
01:12Chunking
Chunking is a powerful cognitive technique that improves short-term memory retention by organizing information into smaller, more manageable units. The brain, limited by working memory capacity, can more easily process and store information when it is divided into "chunks" rather than presented as discrete, unrelated elements. Chunking is especially useful when dealing with large amounts of information, such as numerical sequences, words, or complex ideas.
The principle behind chunking is...
The principle behind chunking is...
01:20Introduction to Cognitive Psychology
Cognitive psychology is the field of psychology dedicated to examining how people think. It attempts to explain how and why we think the way we do by studying the interactions among human thinking, emotion, creativity, language, and problem-solving, as well as other cognitive processes. Cognitive psychology studies how information is processed and manipulated in remembering, thinking, and knowing.
This field emerged in the mid-20th century, following a period dominated by behaviorism, which...
This field emerged in the mid-20th century, following a period dominated by behaviorism, which...
01:17Cognitivism
Cognitive psychology emerged as a significant field in the mid-20th century. It focused on understanding humans' internal mental processes. This approach emphasizes how people perceive, remember, think, and solve problems—elements critical to human cognition.
Previously dominated by behaviorism, which prioritized observable behaviors and largely ignored mental processes, psychology transformed in the 1950s. Cognitive psychologists argue that understanding how we think and process information is...
Previously dominated by behaviorism, which prioritized observable behaviors and largely ignored mental processes, psychology transformed in the 1950s. Cognitive psychologists argue that understanding how we think and process information is...