{
  "schema_version": "1.1",
  "id": "s13:https://arxiv.org/abs/2610.32933v1",
  "slug": "2610-32933v1-0kbcu6k",
  "url": "https://feed7.dev/p/2610-32933v1-0kbcu6k",
  "title": "Yalda-Embedding: Open Development of a Competitive Persian Retrieval Model",
  "why_included": "Yalda Embedding offers a compact Persian retrieval model with released weights, data and code. Its results suggest distillation can improve retrieval without increasing inference size.",
  "summary": "Yalda Embedding adapts a **595.8M-parameter** model with over **1 million query-document pairs**. On eight FaMTEB tasks, its macro-average is **0.5488**, the highest among evaluated 0.6B models.",
  "practical_implication": "Builders adding Persian retrieval can evaluate the released model and inspect the dataset and training code. The final distillation step raised the mean from 0.4813 to 0.5488 without increasing inference-time model size.",
  "agent_context": "Yalda Embedding adapts a **595.8M-parameter** model with over **1 million query-document pairs**. On eight FaMTEB tasks, its macro-average is **0.5488**, the highest among evaluated 0.6B models.\n\nBuilders adding Persian retrieval can evaluate the released model and inspect the dataset and training code. The final distillation step raised the mean from 0.4813 to 0.5488 without increasing inference-time model size.\n\nResults are specific to the reported FaMTEB tasks and one-epoch setup. The paper reports comparisons against selected models, not broader production performance.",
  "source": {
    "name": "arXiv",
    "url": "https://arxiv.org/abs/2610.32933v1",
    "published_at": "2026-10-23T17:30:40.000Z"
  },
  "source_class": "blog_post",
  "content_type": "Paper",
  "layer": "model",
  "domains": [
    "research",
    "data"
  ],
  "topics": [
    "retrieval",
    "open-models"
  ],
  "verification": {
    "status": "needs_review",
    "label": "Needs Review",
    "method": "unverified",
    "verified_at": null
  },
  "uncertainty": [
    "Results are specific to the reported FaMTEB tasks and one-epoch setup. The paper reports comparisons against selected models, not broader production performance."
  ],
  "connected_context": {
    "meaning": "This adds Persian retrieval evidence for a compact open model, with the reported distillation gain improving benchmark performance without increasing inference-time size. It makes language-specific retrieval a concrete model option to evaluate, while keeping the case bounded to the reported FaMTEB tasks and setup.",
    "corpus_size": 910,
    "generated_at": "2026-10-26T11:01:08.586Z",
    "connections": [
      {
        "title": "superlinked/sie",
        "source_name": "GitHub",
        "source_url": "https://github.com/superlinked/sie",
        "feed7_url": "https://feed7.dev/p/sie-055bka3",
        "reason": "Yalda supplies a specialized retrieval model that could fit the kinds of task-specific models SIE serves behind a unified API; the paper does not evaluate integration or serving through SIE."
      },
      {
        "title": "When Does Bigger Help? A Controlled Study of LLM Scale for Ontology Learning",
        "source_name": "arXiv",
        "source_url": "https://arxiv.org/abs/2608.31118v1",
        "feed7_url": "https://feed7.dev/p/2608-31118v1-1abyqwd",
        "reason": "Both report that model choice should be judged on the target task: Yalda’s result is specific to Persian retrieval benchmarks, while the ontology study finds size rankings vary by task."
      }
    ]
  },
  "lifecycle": "Current",
  "published_at": "2026-10-23T17:30:40.000Z",
  "modified_at": "2026-10-23T17:30:40.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/2610-32933v1-0kbcu6k",
    "json": "https://feed7.dev/p/2610-32933v1-0kbcu6k.json",
    "markdown": "https://feed7.dev/p/2610-32933v1-0kbcu6k.md"
  }
}