> ## Documentation Index
> Fetch the complete documentation index at: https://docs.amlok.io/llms.txt
> Use this file to discover all available pages before exploring further.

# References and source scope

> Original data publishers, benchmark definitions, evaluation methods and governance sources supporting AMLOK's public research documentation.

Use these references to trace AMLOK's public data, benchmark definitions, evaluation methods and governance requirements. Each source retains its own scope and use conditions. A citation alone does not establish permission to train, independent replication or an AMLOK result. Record the exact revision used in each experiment.

## Method definitions

| Source | Role in this program |
| - | - |
| [LoRA](https://arxiv.org/abs/2106.09685) | Parameter-efficient adaptation method. |
| [Ditto: Deep Entity Matching with Pre-Trained Language Models](https://arxiv.org/abs/2004.00584) | Learned entity-matching baseline and experimental context. |

## Evaluation and baselines

| Source | Role in this program |
| - | - |
| [FinBen](https://arxiv.org/abs/2402.12659) | Broad financial task taxonomy; no substitute for Canadian lending labels. |
| [FinanceBench](https://arxiv.org/abs/2311.11944) | Financial-document question-answering evaluation. |
| [FinQA](https://arxiv.org/abs/2109.00122) | Numerical reasoning with financial evidence. |
| [TAT-QA](https://arxiv.org/abs/2105.07624) | Reasoning over financial text and tables. |
| [Revisiting Deep Learning Models for Tabular Data](https://arxiv.org/abs/2106.11959) | FT-Transformer and strong tabular baselines. |
| [TabPFN 3.5 technical report](https://priorlabs.ai/technical-reports/tabpfn-3-5) | Current tabular-model candidate; access and permitted use require checking for an actual run. |
| [On Calibration of Modern Neural Networks](https://arxiv.org/abs/1706.04599) | Probability calibration and validation methodology. |
| [OpenAI model announcement](https://openai.com/index/gpt-6-astra/), [Anthropic model page](https://www.anthropic.com/claude/fable), [Google Gemini catalogue](https://deepmind.google/models/gemini/) | Sources for candidate model availability. No hosted frontier model was evaluated in this package. |

## Canadian sources and governance

| Source | Role in this program |
| - | - |
| [Corporations Canada data services](https://ised-isde.canada.ca/site/corporations-canada/en/data-services) and [API documentation](https://api.ised-isde.canada.ca/en/docs?api=corporations\&wbdisable=true) | Federal registry scope and access mechanism. |
| [Office of the Superintendent of Bankruptcy statistics](https://ised-isde.canada.ca/site/office-superintendent-bankruptcy/en/statistics-and-research) | Aggregate insolvency context, not individual labels. |
| [Statistics Canada WDS](https://www.statcan.gc.ca/en/developers/wds/user-guide) and [open licence guidance](https://www.statcan.gc.ca/en/terms-conditions/open-licence-faq) | Public statistical acquisition and reuse context. |
| [Bank of Canada Valet documentation](https://www.bankofcanada.ca/valet/docs) and [terms](https://www.bankofcanada.ca/terms/) | Source for the executed FX descriptive analysis. |
| [BDC debt-service coverage definition](https://www.bdc.ca/en/articles-tools/entrepreneur-toolkit/templates-business-guides/glossary/debt-service-coverage-ratio) | Financial definition reference; agreement-specific ratios can differ. |
| [CMHC GDS and TDS calculation](https://www.cmhc-schl.gc.ca/professionals/project-funding-and-mortgage-financing/mortgage-loan-insurance/calculating-gds-tds) | Mortgage affordability definitions within CMHC's stated scope. |
| [OSFI Guideline E-23 (2027)](https://www.osfi-bsif.gc.ca/en/guidance/guidance-library/guideline-e-23-model-risk-management-2027) | Model-risk governance; revised guideline effective May 1, 2027. |
| [Canadian privacy authorities' generative-AI principles](https://www.priv.gc.ca/en/privacy-topics/technology/artificial-intelligence/gd_principles_ai/) | Privacy, purpose, and accountability reference. |
| [Québec private-sector privacy statute, section 12.1](https://www.legisquebec.gouv.qc.ca/fr/version/lc/P-39.1?code=se%3A12_1\&history=20250414\&langCont=en) | Exclusively automated decisions within the provision's scope; linked official version must be rechecked for an actual deployment. |

The initial research consulted public source pages and selected AMLOK starter documents. It did not acquire competitor datasets, commercial model weights, private borrower histories, or a Canadian loan-outcome corpus. Public-data availability and source quality are evaluated separately from permission to collect, redistribute, or train.

## Original records and acquisition sources

The [public-source inventory](https://docs.amlok.io/research/canadian-data.md) contains the full source-by-source access and field register. The executed source analysis uses original [EDC disclosures](https://www.edc.ca/en/about-us/corporate/disclosure/reporting-transactions/individual-transaction-information.html), [Investissement Québec intervention records](https://www.investquebec.com/fr/tout-sur-nous/diffusion-de-linformation/archives), [CSBFP data](https://open.canada.ca/data/dataset/3f718801-099d-4037-bb0a-1d41ba8aca8b), [Kinross financing disclosure](https://www.kinross.com/news-and-investors/news-releases/press-release-details/2022/Kinross-announces-US1.0-billion-term-loan/default.aspx), [Lundin Mining's amendment](https://www.lundinmining.com/news/lundin-mining-increases-existing-credit-facility-to-us45-billion), and [BCE's financial report](https://www.bce.ca/cs-assets/2025-bce-annual-financial-report.pdf-blt5d552f866d8a58a3).

The [SBA data portal](https://data.sba.gov/dataset/7a-504-foia) supplies the inspected US loan schema. The original [FinQA repository](https://github.com/czyssrs/FinQA) supplies the executed retrieval task; [TAT-QA](https://github.com/NExTplusplus/TAT-QA) supplies development examples. [FinanceBench's publisher page](https://www.patronus.ai/research/financebench) identifies the open sample and its use conditions. [CUAD](https://www.atticusprojectai.org/cuad) is a candidate contract-clause benchmark, not an executed result here.

The [acquisition manifest](https://docs.amlok.io/downloads/research/public-records/acquisition-manifest.json) records the exact file URLs and hashes. A citation alone is not evidence that the entire archive was collected, labels were adjudicated or a model was trained.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.