Triple-Entry Accounting as a Means of Auditing Large Language Models
Bibliographic record
Abstract
The usage of Large Language Models (LMMs) and their exponential progress has created a Cambrian Explosion in the development of new tools for almost every field of science and technology, but also presented significant concerns regarding the AI ethics and creation of sophisticated malware and phishing attacks. Moreover, several worries have arisen in the field of dataset collection and intellectual property in that many datasets may exist without the license of the respective owners. Triple-Entry Accounting (TEA) has been proposed by Ian Grigg to increase transparency, accountability, and security in financial transactions. This method expands upon the traditional double-entry accounting system, which records transactions as debits and credits in two separate ledgers, by incorporating a third ledger as an independent verifier via a digitally signed receipt. The utilization of a digital signature provides evidentiary power to the receipt, thus reducing the accounting problem to one of the presence or absence of the receipt. The integrity issues associated with double-entry accounting can be addressed by allowing the parties involved in the transaction to share the records with an external auditor. This manuscript proposes a novel methodology to apply triple-entry accounting records on a publicly accessed distributed ledger technology medium to control the queries of LLMs in order to discourage malicious acts and ensure intellectual property rights.
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 | 0.007 | 0.034 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.006 | 0.009 |
| Open science | 0.002 | 0.004 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
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".