{
  "schema_version": "1.1",
  "id": "s13:https://arxiv.org/abs/2610.37651v1",
  "slug": "2610-37651v1-1usz0jw",
  "url": "https://feed7.dev/p/2610-37651v1-1usz0jw",
  "title": "GEPA-Curator: Curated Learning for Self Improving Agents",
  "why_included": "GEPA-Curator filters noisy agent failures and retrieves relevant successful runs before prompt optimization, improving results over two baselines across four environments.",
  "summary": "GEPA-Curator adds **task curation** to filter experiences with unreliable learning signals and **prior success retrieval** to add relevant successful trajectories as context. It was evaluated on four environments with Gemini 3.1 Pro and Claude Opus 5 as optimizer models.",
  "practical_implication": "If you optimize prompts from coding-agent runs, separate useful failures from harness issues, stochastic outcomes, and noisy judges; consider reusing successful trajectories for similar tasks.",
  "agent_context": "GEPA-Curator adds **task curation** to filter experiences with unreliable learning signals and **prior success retrieval** to add relevant successful trajectories as context. It was evaluated on four environments with Gemini 3.1 Pro and Claude Opus 5 as optimizer models.\n\nIf you optimize prompts from coding-agent runs, separate useful failures from harness issues, stochastic outcomes, and noisy judges; consider reusing successful trajectories for similar tasks.\n\nThe paper reports gains of up to **20.7% over GEPA** and **21.2% over ACE**, but the material does not specify how these gains vary by environment or how much curation costs.",
  "source": {
    "name": "arXiv",
    "url": "https://arxiv.org/abs/2610.37651v1",
    "published_at": "2026-10-27T17:58:11.000Z"
  },
  "source_class": "blog_post",
  "content_type": "Paper",
  "layer": "agent",
  "domains": [
    "research"
  ],
  "topics": [
    "harness-engineering",
    "prompting",
    "agent-evals"
  ],
  "verification": {
    "status": "needs_review",
    "label": "Needs Review",
    "method": "unverified",
    "verified_at": null
  },
  "uncertainty": [
    "The paper reports gains of up to **20.7% over GEPA** and **21.2% over ACE**, but the material does not specify how these gains vary by environment or how much curation costs."
  ],
  "connected_context": {
    "meaning": "GEPA-Curator adds two controls to prompt optimization: filter experiences with unreliable learning signals and retrieve relevant successful trajectories. The reported gains are bounded by four environments and two optimizer models; the supplied record leaves variation by environment and curation cost unspecified.",
    "corpus_size": 944,
    "generated_at": "2026-10-28T11:08:09.163Z",
    "connections": [
      {
        "title": "Improving Agents is a Data Mining Problem — Vivek Trivedy, LangChain",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=CvRngaQZQ3Y",
        "feed7_url": "https://feed7.dev/p/improving-agents-is-a-data-mining-problem-vivek-trivedy-langchain-1cgibr8",
        "reason": "LangChain describes mining production traces for failures and replayable evals; GEPA-Curator adds a way to screen those experiences for usable learning signals and retrieve relevant successes."
      },
      {
        "title": "The Rise of Verbal Reinforcement Learning",
        "source_name": "arXiv",
        "source_url": "https://arxiv.org/abs/2609.01597v1",
        "feed7_url": "https://feed7.dev/p/2609-01597v1-1yaem2a",
        "reason": "The survey distinguishes feedback retained as context from other learning interventions; GEPA-Curator's successful-trajectory retrieval is a context intervention, while its task curation filters optimizer inputs."
      },
      {
        "title": "FOCUS: Active Critic Learning for Long-Horizon Language Model Agents",
        "source_name": "arXiv",
        "source_url": "https://arxiv.org/abs/2610.35801v1",
        "feed7_url": "https://feed7.dev/p/2610-35801v1-1bbefst",
        "reason": "FOCUS selectively samples states to improve long-horizon critic estimates, while GEPA-Curator filters whole experiences for reliable learning signals; both address where optimization effort should be spent, at different stages."
      }
    ]
  },
  "lifecycle": "Current",
  "published_at": "2026-10-27T17:58:11.000Z",
  "modified_at": "2026-10-27T17:58:11.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/2610-37651v1-1usz0jw",
    "json": "https://feed7.dev/p/2610-37651v1-1usz0jw.json",
    "markdown": "https://feed7.dev/p/2610-37651v1-1usz0jw.md"
  }
}