{
  "schema_version": "1.1",
  "id": "archive:https://arxiv.org/abs/2608.18033v1",
  "slug": "2608-18033v1-0gvjv1y",
  "url": "https://feed7.dev/p/2608-18033v1-0gvjv1y",
  "title": "Where A Small Language Model Helps in Invoice Categorisation, Understood Through Embedding Geometry",
  "why_included": "A single-GPU SBERT beat the reported zero-shot LLM and vendor baseline for invoice coding, suggesting narrow, private classifiers can outperform broader models with modest local data.",
  "summary": "A fine-tuned SBERT reached **0.96 accuracy** on invoice classification, above the study’s zero-shot LLM and vendor baseline. For new-client generalization it reached **0.9 F1** with roughly **100 client-specific invoices**, using one GPU.",
  "practical_implication": "For narrow classification with sensitive data, benchmark a small in-house encoder before defaulting to a hosted general model. Inspect embedding clusters and test raw versus human-friendly structured inputs rather than assuming extra formatting helps.",
  "agent_context": "A fine-tuned SBERT reached **0.96 accuracy** on invoice classification, above the study’s zero-shot LLM and vendor baseline. For new-client generalization it reached **0.9 F1** with roughly **100 client-specific invoices**, using one GPU.\n\nFor narrow classification with sensitive data, benchmark a small in-house encoder before defaulting to a hosted general model. Inspect embedding clusters and test raw versus human-friendly structured inputs rather than assuming extra formatting helps.\n\nThe results concern one financial corpus and task, so they do not establish a general small-model advantage. Vendor identity strongly shaped the embedding space, and the material does not report deployment cost or performance on other accounting datasets.",
  "source": {
    "name": "arXiv",
    "url": "https://arxiv.org/abs/2608.18033v1",
    "published_at": "2026-08-18T17:28:01.000Z"
  },
  "source_class": "blog_post",
  "content_type": "Paper",
  "layer": "model",
  "domains": [
    "data"
  ],
  "topics": [
    "model-selection"
  ],
  "verification": {
    "status": "needs_review",
    "label": "Needs Review",
    "method": "unverified",
    "verified_at": null
  },
  "uncertainty": [
    "The results concern one financial corpus and task, so they do not establish a general small-model advantage. Vendor identity strongly shaped the embedding space, and the material does not report deployment cost or performance on other accounting datasets."
  ],
  "connected_context": null,
  "lifecycle": "Current",
  "published_at": "2026-08-18T17:28:01.000Z",
  "modified_at": "2026-08-18T17:28:01.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/2608-18033v1-0gvjv1y",
    "json": "https://feed7.dev/p/2608-18033v1-0gvjv1y.json",
    "markdown": "https://feed7.dev/p/2608-18033v1-0gvjv1y.md"
  }
}