An AI Multi-Model Approach to DeFi Project Trust Scoring and Security
Bibliographic record
Abstract
Rampant scams plague decentralized finance (DeFi) projects, creating a DeFi credibility problem that limits the impact of DeFi advances in the availability and variety of financial services. This paper presents a novel solution to the DeFi credibility problem by developing an AI multi-model that generates TrustScore ratings for DeFi projects and clear explanations of the scores. We generate DeFi-project TrustScore by aggregating multiple factors that provide DeFi investors with a holistic view of DeFi project trustworthiness. To rate a DeFi project with a TrustS core, we combine the output of four AI pipelines that analyze smart contract code vulnerabilities, suspicious transactions, anomalous price changes to smart contracts, and social media scam sentiment. Applying four factors exponentially improves the trust-score accuracy over the single-factor approaches done historically. Two of the factors, anomalous price change, and social media sentiment, have not been used before to detect DeFi fraud. Furthermore, we enhanced the most critical factor, smart-contract code vulnerability detection, with the latest Large Language Models (LLMs). Our overall system is a multi-model composed of a TrustS core Explainer LLM that aggregates individual pipeline results, a fine-tuned GPT model to audit smart contract code, the Prophet forecasting tool, FinBERT tailored for financial Natural Language Processing (NLP), and XGBoost for classification. The proposed approach identifies a significant proportion of known fraudulent DeFi projects and generates an accurate and explained TrustScore. Thus, we address the DeFi credibility problem so that investors can make reliable decisions about DeFi projects.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.009 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".