AIMS-QA: SentenceBERT retriever model

Retrieval model accompanying AIMS-QA (Data & Policy, 2026), a multi-agent framework for assessing corporate modern slavery statements through question answering rather than binary compliance checking. FILES sentencebert_retriever.zip (80 MB) — Fine-tuned SentenceBERT retriever used by the AIMS-QA RAG Watchdog (Agent 2) for question-level evidence retrieval. Contains the complete SentenceTransformer directory: model weights, pooling configuration and tokenizer. Unzip and load with SentenceTransformer("sentencebert_retriever"). Base model: sentence-transformers/all-MiniLM-L6-v2. WEIGHTS NOT INCLUDED HERE Agent 1, the evidence-extraction classifier, uses the AIMSDistill Australian student (au.pth, 1.47 GB). It is not duplicated in this record — see doi:10.6084/m9.figshare.33261576 The ModernBERT backbone is the unmodified public model answerdotai/ModernBERT-large, downloaded automatically by the transformers library. USEQuestion-level retrieval of evidence from corporate modern slavery statements, against the eleven mandatory reporting criteria of the Australian Modern Slavery Act 2018, as defined in AIMS.au paper.The question set used by the framework is available in the code repository at papers/aimsqa/src/data/questions.csv. The full 209-question taxonomy developed for AIMS-QA is not distributed; researchers who wish to use it should contact the corresponding author.These models are decision support for human reviewers, not automated compliance verdicts. Misclassifying a corporate disclosure carries legal and ethical consequences; the paper states that full automation is not recommended.CHECKSUMS (SHA-256) model.safetensors 2afcbb92742e8fd98f5cdfdd41dc77f33fcba34b54fdd0fd4d7347c8a67a7660 CITE Bora AE, Zhang D, Thinyane H, Bashar MA, Nayak R, Mengersen K (2026). AIMS-QA: A Framework for Scaling Beyond Compliance the Assessment of Corporate Modern Slavery Statements with AI-driven Question Answering. Data & Policy, Cambridge University Press. Code: https://github.com/mila-studios/ai4h_aims-au