| Proj No. | A3243-251 |
| Title | Large Language Model: Methods to alleviate hallucinations |
| Summary | Large Language Model is a form of large neural network that handles Natural Language Processes. Neural networks such as LLM suffers from hallucinations, where the LLM invents new information to fill in the gaps in the knowledge. This is due to the fact that neural networks are, by design, a predictor. However, this new information tends to be incorrect or have no logical reasoning. There are many newly proposed and implemented frameworks and fixes that attempt to mitigate hallucinations by improving user inputs or evaluate how the LLM approaches problems. The aim of this project is to create a framework reduce hallucinations experienced by the LLM. The goal is to increase the quality and correctness of the responses provided to the users. This can benefit anyone that uses LLMs or fields that interact with LLMs, such as robotics where it can improve orders given to a robot. |
| Supervisor | A/P Andy Khong Wai Hoong (Loc:Admin Building, Ext: +65 67906008) |
| Co-Supervisor | - |
| RI Co-Supervisor | - |
| Lab | Delta-NTU Copr Lab (Loc: S2-B3c-13 (Tel: 6592 2661)) |
| Single/Group: | Single |
| Area: | Digital Media Processing and Computer Engineering |
| ISP/RI/SMP/SCP?: |