{
  "schema_version": "1.1",
  "id": "s13:https://arxiv.org/abs/2610.31667v1",
  "slug": "2610-31667v1-1jz3l6f",
  "url": "https://feed7.dev/p/2610-31667v1-1jz3l6f",
  "title": "Direct Pass@$k$ Optimization for Thought Pruning: Balancing Quality and Diversity in Test-Time Reasoning",
  "why_included": "A pruning objective selects reasoning branches for their combined pass@k, retaining promising but nonredundant paths so test-time compute is spent on a stronger candidate pool.",
  "summary": "The method scores a pool of partial solutions by the chance that at least one continuation will be correct. It was tested on **eight open-weight reasoning models** and outperformed quality-only and diversity-only pruning.",
  "practical_implication": "For agent workloads that sample several reasoning paths, optimize the surviving set rather than ranking branches independently. **Direct pass@k optimization** makes both predicted quality and overlap between candidates part of one selection criterion.",
  "agent_context": "The method scores a pool of partial solutions by the chance that at least one continuation will be correct. It was tested on **eight open-weight reasoning models** and outperformed quality-only and diversity-only pruning.\n\nFor agent workloads that sample several reasoning paths, optimize the surviving set rather than ranking branches independently. **Direct pass@k optimization** makes both predicted quality and overlap between candidates part of one selection criterion.\n\nThe abstract provides no effect sizes or compute-cost comparison. Gains were strongest with a **moderate number of survivors**, so applicability to very small or very large candidate pools remains unclear.",
  "source": {
    "name": "arXiv",
    "url": "https://arxiv.org/abs/2610.31667v1",
    "published_at": "2026-10-22T17:59:48.000Z"
  },
  "source_class": "blog_post",
  "content_type": "Paper",
  "layer": "benchmark",
  "domains": [
    "research"
  ],
  "topics": [
    "reasoning",
    "agent-evals"
  ],
  "verification": {
    "status": "needs_review",
    "label": "Needs Review",
    "method": "unverified",
    "verified_at": null
  },
  "uncertainty": [
    "The abstract provides no effect sizes or compute-cost comparison. Gains were strongest with a **moderate number of survivors**, so applicability to very small or very large candidate pools remains unclear."
  ],
  "connected_context": {
    "meaning": "This narrows test-time branch pruning to a set-level objective: retain candidates according to the probability that at least one continuation succeeds, rather than treating quality and diversity as separate heuristics. It supports evaluating reasoning as a multi-path inference system, but absent effect sizes and compute comparisons leave the practical advantage and useful pool sizes unresolved.",
    "corpus_size": 898,
    "generated_at": "2026-10-24T10:28:17.103Z",
    "connections": [
      {
        "title": "Test-Time Scaling in Reasoning LLMs: Inference Regimes, Evaluation, and Reproducibility",
        "source_name": "arXiv",
        "source_url": "https://arxiv.org/abs/2608.04001v1",
        "feed7_url": "https://feed7.dev/p/2608-04001v1-1gj91hk",
        "reason": "Direct pass@k pruning instantiates one specific multi-path inference regime, reinforcing the need to report survivor counts, sampling and pruning protocol, compute, and uncertainty rather than attributing gains to the base model alone."
      },
      {
        "title": "Legibility is Not Interpretability: Comparing Judged and Actual Importance in Chain-Of-Thought Reasoning",
        "source_name": "arXiv",
        "source_url": "https://arxiv.org/abs/2609.04194v1",
        "feed7_url": "https://feed7.dev/p/2609-04194v1-0milybu",
        "reason": "The set-level success objective avoids relying solely on readable or independently judged reasoning steps, consistent with evidence that apparent step importance need not reflect causal contribution."
      }
    ]
  },
  "lifecycle": "Current",
  "published_at": "2026-10-22T17:59:48.000Z",
  "modified_at": "2026-10-22T17:59:48.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/2610-31667v1-1jz3l6f",
    "json": "https://feed7.dev/p/2610-31667v1-1jz3l6f.json",
    "markdown": "https://feed7.dev/p/2610-31667v1-1jz3l6f.md"
  }
}