{ "schema_version": 1, "builder_plan": { "scenario": "A multi-step product task fails because the agent loses track of dependencies.", "change": "Try the proposed planning step on tasks with checkable intermediate outcomes.", "baseline": "The same agent solving the same tasks without the new planning step.", "steps": [ "Locate the paper's implementation and check its license, model support, data needs, and hardware requirements. Stop here if you cannot obtain the required artifacts.", "Save a small set of representative product tasks and expected outcomes. Keep evaluation tasks out of any training or tuning, and choose quality and cost limits before running.", "Run the baseline and the proposed change on identical inputs. Keep other settings fixed and repeat paired runs to expose variation.", "Save per-task outputs, failures, timings, and all usage costs. Review regressions before expanding the test." ], "measure": "Task completion, unnecessary steps, planning tokens, response time, and total cost.", "decision": "Keep the change only if it meets your preselected quality and latency limits and improves the metric you care about after setup and operating costs. Otherwise retain the baseline. This is an exploration plan, not a validated recommendation." }, "builder_review": null, "arxiv_id": "2609.39069", "title": "CORE: Conflict-Oriented Reasoning Elimination for Verifiable Language-Model Search", "summary": "CORE replaces chronological backtracking and restart-style repair in test-time LM search with verifier-certified conflict cores that direct backjumps to the causally responsible decision and cache conflicts to prevent repetition. The cost is integration complexity: a verifier capable of producing certified conflict cores, conflict storage, and a search controller, plus the proposal budget of systematic search rather than single-pass generation. It can fail when no sound verifier exists for the task, when verifiers cannot localize failures to a decision core, when the search is capped so completeness guarantees do not hold, or on open-ended generation where gains were demonstrated only on structured reasoning tasks with exact verification.", "rationale": "The method is implementable with a single 24 GB GPU using 7-8B models and reduces token and verifier cost, but its benefits depend on having a task-specific sound verifier, so the reader should validate it on workloads where such verification is available before adopting it broadly.", "technique": "Conflict-directed backjumping for LM search", "family": "planejamento_e_decomposicao", "recommendation": "testar", "published": "2026-09-30", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2609.39069", "independent_implementations": 0, "total_implementations": 0, "stars": 0, "citations": 0, "claims": [ { "claim_id": "claim-01", "statement": "CORE replaces chronological backtracking and restart-style repair in test-time LM search with verifier-certified conflict cores that direct backjumps to the causally responsible decision and cache conflicts to prevent repetition. The cost is integration complexity: a verifier capable of producing certified conflict cores, conflict storage, and a search controller, plus the proposal budget of systematic search rather than single-pass generation. It can fail when no sound verifier exists for the task, when verifiers cannot localize failures to a decision core, when the search is capped so completeness guarantees do not hold, or on open-ended generation where gains were demonstrated only on structured reasoning tasks with exact verification.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.39069", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "CORE reduces median verifier calls by 39.8% (30 variables) and 35.0% (36 variables) versus chronological repair on planted graph-coloring instances; on five reasoning tasks it reaches 75.9% mean success with Qwen2.5-7B-Instruct and 84.2% with Qwen3-8B, versus 72.5% and 81.8% for Tree of Thoughts, while using fewer verifier calls and generated tokens.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.39069", "source_page": null, "source_excerpt": "" } ], "exposure_map": [ { "dimension": "quality", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "compute", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "latency", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "operations", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "compatibility", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "security", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "data_and_training", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "reproducibility", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" } ], "risks": [], "minimum_test": [], "open_questions": [], "independent_tests": [], "lab_reproductions": [], "report_available": false, "validation_tier": "unknown", "evidence_tier": "unknown", "infrastructure_basis": "unknown", "equations": [], "equations_status": "not_fetched", "core_kind": "", "equations_fetched_at": "", "equations_source_url": "", "equations_source_sha256": "", "exposures_reviewed_at": "", "exposures_source_sha256": "" }