Anti-hallucination filter

Inspiration Will be honest, this project was created by me a long time ago, but every time I used to improve it for global Hakathons. This website was built in the on the foundation of the novel axiomatic framework I established myself.

What it does The function of the proposed system is to eliminate totally hallucinations of LLM models

How we built it As I mentioned in the Inspiration division, the system is built on the new mathematical discipline. Let me describe the axiomatic framework of this discipline: There are 3 global sets: False set, True set and Undefined set. The information is sent to these sets depending on the sign of context operator, which works with contextual quantitative data and sources. There are 3 context operators for each set. K+ (True set), K- (False set), K_0 (Undefined set). For example: If K=K+, S (information)€T. The next thing I want to talk about is their constraints: K+€+Z; K-€-Z;

K_0=0. Now let me switch to the dynamic side of my system: K_0+Delta_K=K_1. This formula denotes that the information which had a certain amount of context sources in the future can change by Delta_K

Challenges we ran into The only challenge was exactly about how to make my system dynamical, but as you saw in the previous division I handled it

Accomplishments that we're proud of To be honest, I'm really proud of what I have created so far and what results I have achieved in understanding websites and AI

What we learned What I want to say is that I was really passionate about what I have been doing. During this process, I have learned how to make a deal with AI modules, how to create websites and how to work within a deadline

What's next for Anti-hallucination filter. Actually, one of the next and main steps after Anti-hallucination filter is to focus on another projects that I left for future

Built With

• api
• claude
• css
• html
• javascript
• json
• python
• wikipedia