Terms of Service

Welcome to candyscript.club. These Terms of Service ("Terms") constitute a legally binding agreement between you ("User," "Researcher," or "You") and CandyScript Research Lab (operated by CandyScript Club Inc., "we," "us," or "our"). By accessing our portal, downloading scientific materials, or using our interactive benchmark calculator, you signify your unreserved assent to these Terms.

1. Nature of the Service & Free Distribution Charter

CandyScript Research Lab is an independent scientific research initiative dedicated to the study of autonomous conversational artificial intelligence, multi-modal turn-taking latency, and 4D facial micro-expression synthesis.

All research whitepapers, evaluation datasets, and interactive benchmark tools provided on this portal are 100% free of charge. There are no recurring membership fees, premium paywalls, or hidden financial obligations. Access to our publications is granted as an academic and technological resource for legitimate educational, scientific, and industry evaluation purposes in exchange for verified contact data.

2. Intellectual Property & Academic Fair Use

All published research manuscripts, benchmark schematics, tensor architectures, and source code excerpts published on candyscript.club are the intellectual property of CandyScript Research Lab and are protected by United States and international copyright statutes.

Unless otherwise designated, downloaded research publications are licensed under the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) license. You are permitted to share, cite, and adapt the material provided that proper academic attribution is accorded to CandyScript Research Lab. Commercial resale or bundling of raw benchmark repositories without express written consent is strictly prohibited.

3. Prohibited Conduct & Ethical AI Use Policy

You agree not to utilize any benchmarks, facial blendshape vectors, or models obtained through this site for:

• Generating unauthorized, non-consensual biometric deepfakes or impersonation of private individuals.
• Developing automated deceptive conversational agents designed to perpetrate financial fraud, social engineering, or voter manipulation.
• Conducting reverse-engineering attacks intended to bypass model safety filters or extract proprietary training corpus data.
• Deploying high-frequency automated scraping scripts that overload our repository infrastructure.

4. Disclaimer of Scientific Warranties

The materials, benchmark predictions, and mathematical formulas are provided on an "as-is" and "as-available" basis for experimental and research evaluation purposes only. While our laboratory strives for rigorous peer-reviewed accuracy, CandyScript Research Lab expressly disclaims all warranties of any kind, whether express, statutory, or implied, including fitness for a particular commercial application.

5. Limitation of Liability

In no event shall CandyScript Research Lab, CandyScript Club Inc., its directors, researchers, or affiliates be liable for any indirect, consequential, exemplary, or punitive damages arising from your access to, inability to access, or implementation of algorithmic parameters derived from our whitepapers or calculators.

6. Accuracy of User Information

When submitting our research access request forms, you represent and warrant that the provided full name, email address, and direct contact details belong to you or your designated institutional representative. Falsifying identities or utilizing temporary throwaway addresses may result in termination of dataset access credentials.

7. Governing Law and Forum

These Terms and any dispute arising out of or related to them shall be governed exclusively by the laws of the State of California, United States, without regard to conflict of law principles. Any legal action shall be adjudicated in the state or federal courts located within the City and County of San Francisco, California.

8. Contact & Legal Notices

For official inquiries regarding these Terms, academic citations, or institutional permissions, please contact our legal office: