hyungi cd33ded7a8 docs(search): passage-RAG go/no-go = NO-GO (hier evidence 동등, diagnose c4+c5) ...

PR-DocSrv-Hier-PassageRAG-Diagnose-1 c4+c5. 조건부 N=12(retrieval 통제) blind pairwise
(hypothesis-blind subagent, 익명 3-file split). 결과 4-way 수렴 = 동등:
pairwise prehier4/hier3/tie5(no edge) + axis ±0.08 + objective 동일(halluc36/36) +
variance~0(byte-identical 재생성). verbosity artifact 없음(prehier 더 길었으나 승+1).
=> NO-GO: hier-leaf evidence 무이득. hier leaf = section-outline UI 전용 완전 확정
(UI yes / doc-search NO-GO / passage-RAG NO-GO 3영역 종결). 2026-06-21 freeze input only.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

2026-05-25 07:02:46 +00:00

..

baseline_2026-04-07_summary.md

…

baseline_2026-04-07.csv

…

baseline_order_unit_2026-04-20.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

freshness_decay_observation_baseline.json

ops(search): PR-RAG-Time-1 1주 후 재측정 PASS

2026-05-12 15:41:40 +09:00

freshness_decay_observation_baseline.md

ops(search): PR-RAG-Time-1 1주 후 재측정 PASS

2026-05-12 15:41:40 +09:00

freshness_decay_observation_week1_postfix.md

ops(search): reranker drift fix 사후 재측정 (postfix observation)

2026-05-13 12:06:20 +09:00

freshness_decay_observation_week1.json

ops(search): PR-RAG-Time-1 1주 후 재측정 PASS

2026-05-12 15:41:40 +09:00

freshness_decay_observation_week1.md

ops(search): PR-RAG-Time-1 1주 후 재측정 PASS

2026-05-12 15:41:40 +09:00

local_research_M1_throughput.md

ops(reports): local research M1/M2/M3 baseline 등록 (2026-05-02)

2026-05-21 07:25:27 +09:00

local_research_M2_mlx_capacity.md

ops(reports): local research M1/M2/M3 baseline 등록 (2026-05-02)

2026-05-21 07:25:27 +09:00

local_research_M3_embedding_throughput.md

ops(reports): local research M1/M2/M3 baseline 등록 (2026-05-02)

2026-05-21 07:25:27 +09:00

passage_rag_capture_2026-05-25.jsonl

feat(search): passage-RAG capture runner + raw JSONL (diagnose c3)

2026-05-25 06:53:11 +00:00

passage_rag_judge_2026-05-25.md

docs(search): passage-RAG go/no-go = NO-GO (hier evidence 동등, diagnose c4+c5)

2026-05-25 07:02:46 +00:00

passage_rag_judge_key_2026-05-25.json

docs(search): passage-RAG go/no-go = NO-GO (hier evidence 동등, diagnose c4+c5)

2026-05-25 07:02:46 +00:00

passage_rag_judge_pairs_2026-05-25.jsonl

docs(search): passage-RAG go/no-go = NO-GO (hier evidence 동등, diagnose c4+c5)

2026-05-25 07:02:46 +00:00

phase1_1a_run2.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase1_1a.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase1_2b_thr15.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase1_2b.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase1_2c_probes10.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase1_2c_probes20.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase1_2c_v2.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase1_2c.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase1_2g_enhanced.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase1_2g.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase1_3_rerank.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase2_1_baseline_regression.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase2_2_multilingual.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase2_2_narrow.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase2_3_soft_filter.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

phase2_final_baseline.csv

…

phase2_final_candidate.csv

…

phase2_final.md

…

phase05_fusion_comparison.md

…

phase05_legacy.csv

…

phase05_rrf_boost.csv

…

phase05_rrf.csv

…

phase_2a_embedding_decision_2026-05-23.md

feat(eval): Phase 2A Diagnose Phase 3+4 — dispatcher + 3 측정 + decision (H3 bge-m3 유지)

2026-05-23 06:55:13 +00:00

phase_2b_reranker_decision_2026-05-23.md

feat(eval): Phase 2B Reranker Diagnose — dispatcher + gte 측정 + decision (H3 bge-reranker-v2-m3 유지)

2026-05-23 08:37:42 +00:00

pr_ds_priority_gate_1_closure.md

docs(search): DS-Mac-mini-26B-Priority-Gate-1 (B-1) closure 보고서

2026-05-17 08:58:38 +09:00

pr_ds_synthesis_timeout_calibration_1_closure.md

docs(search): DS-Synthesis-Timeout-Calibration-1 (B-3) closure 보고서

2026-05-17 08:07:51 +09:00

pr_hermes_docsrv_search_1_closure.md

docs(hermes): PR-Hermes-Docsrv-Search-1 closure 보고서

2026-05-16 20:07:18 +09:00

pr_hermes_maxtokens_followup_1_closure.md

docs(hermes): PR-Hermes-Skill-Curl-Refine-2 (SHIPPED) + MaxTokens-Followup (PARTIAL+REVERTED)

2026-05-17 07:51:02 +09:00

pr_hermes_multiturn_hard_enforcement_1_closure.md

docs(hermes): PR-Hermes-MultiTurn-Hard-Enforcement-1 closure 보고서

2026-05-17 07:35:07 +09:00

pr_hermes_sandbox_env_propagation_1_closure.md

docs(hermes): PR-Hermes-Sandbox-Env-Propagation-1 closure 보고서

2026-05-17 06:37:35 +09:00

pr_hermes_skill_curl_refine_2_closure.md

docs(hermes): PR-Hermes-Skill-Curl-Refine-2 (SHIPPED) + MaxTokens-Followup (PARTIAL+REVERTED)

2026-05-17 07:51:02 +09:00

pr_hermes_skill_polish_1_closure.md

docs(hermes): PR-Hermes-Skill-Polish-1 closure 보고서

2026-05-17 07:13:53 +09:00

pr_hermes_toolcall_adapter_1_closure.md

docs(hermes): PR-Hermes-ToolCall-Adapter-1 closure 보고서

2026-05-16 20:42:34 +09:00

pr_hermes_websearch_1_closure.md

docs(hermes): PR-Hermes-WebSearch-1 closure 보고서

2026-05-16 20:22:43 +09:00

regression_2026-04-20.csv

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00

storage_inventory_2026-05-11.md

…

v0_2_hier_replace_clean_exact_2026-05-25.csv

docs(search): hier vs legacy go/no-go = NO-GO (replace-diagnose c6)

2026-05-25 05:46:14 +00:00

v0_2_hier_replace_prehier_exact_2026-05-25.csv

docs(search): hier vs legacy go/no-go = NO-GO (replace-diagnose c6)

2026-05-25 05:46:14 +00:00

v0_2_hier_replace_raw_exact_2026-05-25.csv

docs(search): hier vs legacy go/no-go = NO-GO (replace-diagnose c6)

2026-05-25 05:46:14 +00:00

v0_2_hier_replace_sanity_2026-05-25.csv

docs(search): hier vs legacy go/no-go = NO-GO (replace-diagnose c6)

2026-05-25 05:46:14 +00:00

v0_2_phase2a_baseline_snapshot_2026-05-23.csv

feat(eval): Phase 2A Diagnose Phase 3+4 — dispatcher + 3 측정 + decision (H3 bge-m3 유지)

2026-05-23 06:55:13 +00:00

v0_2_phase2a_me5_large_inst_2026-05-23.csv

feat(eval): Phase 2A Diagnose Phase 3+4 — dispatcher + 3 측정 + decision (H3 bge-m3 유지)

2026-05-23 06:55:13 +00:00

v0_2_phase2a_snowflake_l_v2_2026-05-23.csv

feat(eval): Phase 2A Diagnose Phase 3+4 — dispatcher + 3 측정 + decision (H3 bge-m3 유지)

2026-05-23 06:55:13 +00:00

v0_2_phase2b_baseline_snapshot_2026-05-23.csv

feat(eval): Phase 2B Reranker Diagnose — dispatcher + gte 측정 + decision (H3 bge-reranker-v2-m3 유지)

2026-05-23 08:37:42 +00:00

v0_2_phase2b_gte_ml_base_2026-05-23.csv

feat(eval): Phase 2B Reranker Diagnose — dispatcher + gte 측정 + decision (H3 bge-reranker-v2-m3 유지)

2026-05-23 08:37:42 +00:00

v0_2_phase2q_baseline_rebaseline_2026-05-24.csv

fix(search): Phase 2Q variants bug fix + Phase 3 3 measurement 박제

2026-05-24 00:51:56 +00:00

v0_2_phase2q_cand_multi_query_macbook_2026-05-24_cold.csv

fix(search): Phase 2Q variants bug fix + Phase 3 3 measurement 박제

2026-05-24 00:51:56 +00:00

v0_2_phase2q_cand_multi_query_macbook_2026-05-24_warm.csv

fix(search): Phase 2Q variants bug fix + Phase 3 3 measurement 박제

2026-05-24 00:51:56 +00:00

v0_2_phase2q_cand_multi_query_macmini_2026-05-24_cold.csv

fix(search): Phase 2Q variants bug fix + Phase 3 3 measurement 박제

2026-05-24 00:51:56 +00:00

v0_2_phase2q_cand_multi_query_macmini_2026-05-24_warm.csv

fix(search): Phase 2Q variants bug fix + Phase 3 3 measurement 박제

2026-05-24 00:51:56 +00:00

v0_2_phase2q_eval_dedup_baseline_2026-05-24.csv

chore(eval): graded NDCG dedup + warning + audit stats (Phase 2Q inflation 정정)

2026-05-24 04:35:33 +00:00

v0_2_phase2q_eval_dedup_gemma_2026-05-24.csv

chore(eval): graded NDCG dedup + warning + audit stats (Phase 2Q inflation 정정)

2026-05-24 04:35:33 +00:00

v0_2_phase2q_rerank_fix_2026-05-24.csv

fix(search): Phase 2Q rerank payload — chunk_id dedup + cap 60 + TEI batch 64 (Apply prereq)

2026-05-24 03:54:59 +00:00

v0_2_phase2q_result_dedup_gemma_2026-05-24.csv

fix(search): Phase 2Q result dedup — apply_diversity unlimited path doc_id inflation 차단

2026-05-24 04:48:50 +00:00

v0_2_phase2q_result_dedup_gemma_warm_2026-05-24.csv

fix(search): Phase 2Q result dedup — apply_diversity unlimited path doc_id inflation 차단

2026-05-24 04:48:50 +00:00

vram_fixture_2026-05-14.md

chore(reports): Phase 1/2 baseline + 2026-04~05 평가·관측 자료 보존

2026-05-15 04:45:56 +00:00