{
  "schema_version": "1.0",
  "id": "s8:https://www.youtube.com/watch?v=hacEQHHhu2Q",
  "slug": "why-large-tiny-lms-agents-on-edge-robotics-cormac-brick-google-0fjif75",
  "url": "https://feed7.dev/p/why-large-tiny-lms-agents-on-edge-robotics-cormac-brick-google-0fjif75",
  "title": "Why Large? Tiny LMs & Agents on Edge/Robotics — Cormac Brick, Google",
  "why_included": "Edge-agent design is often memory-bound: use small prompted models when devices can spare gigabytes, or fine-tune tiny models for narrow functions when reach and latency matter more.",
  "summary": "Google’s optimized **2B model uses 841 MB for weights**, but runtime and KV cache can push it toward **2 GB of active RAM**. A fine-tuned **270M-parameter** model reached **45 tokens/sec and over 86% function-call reliability** across ten mobile actions.",
  "practical_implication": "Start with the device’s memory and latency budget. If a small prompted model fits, use it; otherwise choose a task-specific tiny model, generate representative synthetic training data, and fine-tune it for a narrow function such as voice-to-action.",
  "agent_context": "Google’s optimized **2B model uses 841 MB for weights**, but runtime and KV cache can push it toward **2 GB of active RAM**. A fine-tuned **270M-parameter** model reached **45 tokens/sec and over 86% function-call reliability** across ten mobile actions.\n\nStart with the device’s memory and latency budget. If a small prompted model fits, use it; otherwise choose a task-specific tiny model, generate representative synthetic training data, and fine-tune it for a narrow function such as voice-to-action.\n\nTiny models are not drop-in general agents. Their reliability depends on task-specific data and tuning, while visual workloads and Raspberry Pi interactions can still miss real-time expectations. The reported figures apply to the demonstrated configurations.",
  "source": {
    "name": "AI Engineer",
    "url": "https://www.youtube.com/watch?v=hacEQHHhu2Q",
    "published_at": "2026-07-25T17:00:06.000Z"
  },
  "source_class": "video",
  "content_type": "Video",
  "layer": "model",
  "domains": [],
  "topics": [
    "model-selection",
    "tool-use",
    "open-models"
  ],
  "verification": {
    "status": "source_linked",
    "label": "Source Linked",
    "method": "source_feed",
    "verified_at": null
  },
  "uncertainty": [
    "Tiny models are not drop-in general agents. Their reliability depends on task-specific data and tuning, while visual workloads and Raspberry Pi interactions can still miss real-time expectations. The reported figures apply to the demonstrated configurations."
  ],
  "lifecycle": "Current",
  "published_at": "2026-07-25T17:00:06.000Z",
  "modified_at": "2026-07-25T17:00:06.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/why-large-tiny-lms-agents-on-edge-robotics-cormac-brick-google-0fjif75",
    "json": "https://feed7.dev/p/why-large-tiny-lms-agents-on-edge-robotics-cormac-brick-google-0fjif75.json",
    "markdown": "https://feed7.dev/p/why-large-tiny-lms-agents-on-edge-robotics-cormac-brick-google-0fjif75.md"
  }
}