[{"path":"outputs/beam/hindsight/rag/100k.json","dataset":"beam","run_name":"hindsight","memory":"hindsight","mode":"rag","split":"100k","total_queries":400,"correct":367,"accuracy":0.8618214285714286,"ingestion_time_ms":0.0,"ingested_docs":0,"avg_retrieve_time_ms":2565.3,"avg_context_tokens":23662.0,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/beam/hindsight/rag/1m.json","dataset":"beam","run_name":"hindsight","memory":"hindsight","mode":"rag","split":"1m","total_queries":700,"correct":617,"accuracy":0.7912926675548582,"ingestion_time_ms":2714555.4,"ingested_docs":1228,"avg_retrieve_time_ms":2968.3,"avg_context_tokens":31664.8,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/beam/hindsight/rag/500k.json","dataset":"beam","run_name":"hindsight","memory":"hindsight","mode":"rag","split":"500k","total_queries":700,"correct":610,"accuracy":0.8009276816418098,"ingestion_time_ms":816268.7,"ingested_docs":364,"avg_retrieve_time_ms":1956.0,"avg_context_tokens":30055.3,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/beam/hindsight/single-query/100k.json","dataset":"beam","run_name":"hindsight","memory":"hindsight","mode":"single-query","split":"100k","total_queries":400,"correct":326,"accuracy":0.7336577408538804,"ingestion_time_ms":403580.5,"ingested_docs":6,"avg_retrieve_time_ms":6379.4,"avg_context_tokens":17654.6,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/beam/hindsight/single-query/10m.json","dataset":"beam","run_name":"hindsight","memory":"hindsight","mode":"single-query","split":"10m","total_queries":200,"correct":147,"accuracy":0.6407977206043332,"ingestion_time_ms":5504578.7,"ingested_docs":1,"avg_retrieve_time_ms":14744.9,"avg_context_tokens":27342.0,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/beam/hindsight/single-query/1m.json","dataset":"beam","run_name":"hindsight","memory":"hindsight","mode":"single-query","split":"1m","total_queries":700,"correct":585,"accuracy":0.7386039171368353,"ingestion_time_ms":347733.5,"ingested_docs":51,"avg_retrieve_time_ms":5517.7,"avg_context_tokens":23901.7,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/beam/hindsight/single-query/500k.json","dataset":"beam","run_name":"hindsight","memory":"hindsight","mode":"single-query","split":"500k","total_queries":700,"correct":541,"accuracy":0.7111989151837411,"ingestion_time_ms":21787574.7,"ingested_docs":962,"avg_retrieve_time_ms":5149.3,"avg_context_tokens":20529.5,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/lifebench/hindsight/rag/en.json","dataset":"lifebench","run_name":"hindsight","memory":"hindsight","mode":"rag","split":"en","total_queries":2003,"correct":1433,"accuracy":0.7154268597104343,"ingestion_time_ms":12088621.4,"ingested_docs":2877,"avg_retrieve_time_ms":705.2,"avg_context_tokens":34324.2,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/lifebench/hybrid-search/rag/en.json","dataset":"lifebench","run_name":"hybrid-search","memory":"hybrid-search","mode":"rag","split":"en","total_queries":2003,"correct":1221,"accuracy":0.6095856215676485,"ingestion_time_ms":13975390.8,"ingested_docs":3605,"avg_retrieve_time_ms":791.4,"avg_context_tokens":23544.1,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/locomo/cognee/rag/locomo10.json","dataset":"locomo","run_name":"cognee","memory":"cognee","mode":"rag","split":"locomo10","total_queries":152,"correct":122,"accuracy":0.8026315789473685,"ingestion_time_ms":289744.4,"ingested_docs":19,"avg_retrieve_time_ms":673.7,"avg_context_tokens":14723.5,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/locomo/hybrid-search/rag/locomo10.json","dataset":"locomo","run_name":"hybrid-search","memory":"hybrid-search","mode":"rag","split":"locomo10","total_queries":1540,"correct":1218,"accuracy":0.7909090909090909,"ingestion_time_ms":645739.8,"ingested_docs":272,"avg_retrieve_time_ms":219.3,"avg_context_tokens":22156.5,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/locomo/locomo-hindsight/rag/locomo10.json","dataset":"locomo","run_name":"locomo-hindsight","memory":"hindsight","mode":"rag","split":"locomo10","total_queries":1540,"correct":1417,"accuracy":0.9201298701298701,"ingestion_time_ms":330205.8,"ingested_docs":272,"avg_retrieve_time_ms":964.1,"avg_context_tokens":36235.4,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/longmemeval/hindsight/rag/s.json","dataset":"longmemeval","run_name":"hindsight","memory":"hindsight","mode":"rag","split":"s","total_queries":500,"correct":473,"accuracy":0.946,"ingestion_time_ms":30090034.3,"ingested_docs":11303,"avg_retrieve_time_ms":674.9,"avg_context_tokens":43624.5,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/longmemeval/hybrid-search/rag/s.json","dataset":"longmemeval","run_name":"hybrid-search","memory":"hybrid-search","mode":"rag","split":"s","total_queries":500,"correct":370,"accuracy":0.74,"ingestion_time_ms":57815933.6,"ingested_docs":23867,"avg_retrieve_time_ms":1599.5,"avg_context_tokens":23221.7,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/personamem/cognee/rag/32k.json","dataset":"personamem","run_name":"cognee","memory":"cognee","mode":"rag","split":"32k","total_queries":589,"correct":482,"accuracy":0.8183361629881154,"ingestion_time_ms":2335222.9,"ingested_docs":195,"avg_retrieve_time_ms":657.9,"avg_context_tokens":11848.2,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/personamem/hindsight/rag/32k.json","dataset":"personamem","run_name":"hindsight","memory":"hindsight","mode":"rag","split":"32k","total_queries":589,"correct":510,"accuracy":0.865874363327674,"ingestion_time_ms":1039008.5,"ingested_docs":195,"avg_retrieve_time_ms":674.9,"avg_context_tokens":15811.6,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/personamem/hybrid-search/rag/32k.json","dataset":"personamem","run_name":"hybrid-search","memory":"hybrid-search","mode":"rag","split":"32k","total_queries":589,"correct":497,"accuracy":0.8438030560271647,"ingestion_time_ms":1039209.5,"ingested_docs":195,"avg_retrieve_time_ms":362.2,"avg_context_tokens":24169.0,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null},{"path":"outputs/precisionmembench/hindsight/retrieval/single-turn.json","dataset":"precisionmembench","run_name":"hindsight","memory":"hindsight-cloud","mode":"retrieval","split":"single-turn","total_queries":77,"correct":66,"accuracy":0.8571428571428571,"ingestion_time_ms":15526.4,"ingested_docs":35,"avg_retrieve_time_ms":256.0,"avg_context_tokens":150.5,"category":null,"mean_precision":0.8481,"mean_recall":1.0,"active_passes":33,"active_total":43},{"path":"outputs/sdebench/hindsight-claude-1/coding/boltons.json","dataset":"sdebench","run_name":"hindsight-claude-1","memory":"hindsight-coding","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":0.0,"ingested_docs":0,"avg_retrieve_time_ms":80218.0,"avg_context_tokens":208.7,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"claude-code","model":"claude-sonnet-5","tasks":61,"solved":61,"interventions":24,"cost_usd":21.5,"turns":896,"wall_s":4893.0,"tokens_in":31433379,"tokens_out":277621},{"path":"outputs/sdebench/hindsight-claude-2/coding/boltons.json","dataset":"sdebench","run_name":"hindsight-claude-2","memory":"hindsight-coding","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":0.0,"ingested_docs":0,"avg_retrieve_time_ms":71614.8,"avg_context_tokens":216.1,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"claude-code","model":"claude-sonnet-5","tasks":61,"solved":61,"interventions":17,"cost_usd":19.53,"turns":832,"wall_s":4368.0,"tokens_in":29041651,"tokens_out":251341},{"path":"outputs/sdebench/hindsight-claude-3/coding/boltons.json","dataset":"sdebench","run_name":"hindsight-claude-3","memory":"hindsight-coding","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":0.0,"ingested_docs":0,"avg_retrieve_time_ms":73431.1,"avg_context_tokens":222.6,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"claude-code","model":"claude-sonnet-5","tasks":61,"solved":61,"interventions":25,"cost_usd":20.78,"turns":868,"wall_s":4479.0,"tokens_in":29943843,"tokens_out":255704},{"path":"outputs/sdebench/hindsight-codex-1/coding/boltons.json","dataset":"sdebench","run_name":"hindsight-codex-1","memory":"hindsight-coding","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":0.0,"ingested_docs":0,"avg_retrieve_time_ms":46752.5,"avg_context_tokens":243.0,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"codex","model":"gpt-5.4-mini","tasks":61,"solved":61,"interventions":25,"cost_usd":15.98,"turns":1392,"wall_s":2852.0,"tokens_in":32507346,"tokens_out":465613},{"path":"outputs/sdebench/hindsight-codex-2/coding/boltons.json","dataset":"sdebench","run_name":"hindsight-codex-2","memory":"hindsight-coding","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":0.0,"ingested_docs":0,"avg_retrieve_time_ms":47696.7,"avg_context_tokens":236.5,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"codex","model":"gpt-5.4-mini","tasks":61,"solved":61,"interventions":30,"cost_usd":16.78,"turns":1450,"wall_s":2910.0,"tokens_in":34187227,"tokens_out":484686},{"path":"outputs/sdebench/hindsight-codex-3/coding/boltons.json","dataset":"sdebench","run_name":"hindsight-codex-3","memory":"hindsight-coding","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":0.0,"ingested_docs":0,"avg_retrieve_time_ms":51391.8,"avg_context_tokens":226.2,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"codex","model":"gpt-5.4-mini","tasks":61,"solved":61,"interventions":31,"cost_usd":17.67,"turns":1461,"wall_s":3135.0,"tokens_in":35254187,"tokens_out":567986},{"path":"outputs/sdebench/hindsight-opencode-1/coding/boltons.json","dataset":"sdebench","run_name":"hindsight-opencode-1","memory":"hindsight-coding","mode":"coding","split":"boltons","total_queries":61,"correct":60,"accuracy":0.9836065573770492,"ingestion_time_ms":0.0,"ingested_docs":0,"avg_retrieve_time_ms":171332.8,"avg_context_tokens":249.1,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"opencode","model":"google/gemini-3.5-flash","tasks":61,"solved":60,"interventions":50,"cost_usd":31.77,"turns":2135,"wall_s":10451.0,"tokens_in":54775427,"tokens_out":840575},{"path":"outputs/sdebench/hindsight-opencode-2/coding/boltons.json","dataset":"sdebench","run_name":"hindsight-opencode-2","memory":"hindsight-coding","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":0.0,"ingested_docs":0,"avg_retrieve_time_ms":160829.5,"avg_context_tokens":220.8,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"opencode","model":"google/gemini-3.5-flash","tasks":61,"solved":61,"interventions":46,"cost_usd":32.3,"turns":2115,"wall_s":9811.0,"tokens_in":53326204,"tokens_out":797596},{"path":"outputs/sdebench/hindsight-opencode-3/coding/boltons.json","dataset":"sdebench","run_name":"hindsight-opencode-3","memory":"hindsight-coding","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":0.0,"ingested_docs":0,"avg_retrieve_time_ms":157729.5,"avg_context_tokens":237.3,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"opencode","model":"google/gemini-3.5-flash","tasks":61,"solved":61,"interventions":50,"cost_usd":37.14,"turns":2298,"wall_s":9622.0,"tokens_in":62036828,"tokens_out":849154},{"path":"outputs/sdebench/vanilla-claude-1/coding/boltons.json","dataset":"sdebench","run_name":"vanilla-claude-1","memory":"vanilla","mode":"coding","split":"boltons","total_queries":61,"correct":60,"accuracy":0.9836065573770492,"ingestion_time_ms":13.2,"ingested_docs":14706,"avg_retrieve_time_ms":87980.3,"avg_context_tokens":7.0,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"claude-code","model":"claude-sonnet-5","tasks":61,"solved":60,"interventions":56,"cost_usd":27.29,"turns":1015,"wall_s":5367.0,"tokens_in":36732480,"tokens_out":390756},{"path":"outputs/sdebench/vanilla-claude-2/coding/boltons.json","dataset":"sdebench","run_name":"vanilla-claude-2","memory":"vanilla","mode":"coding","split":"boltons","total_queries":61,"correct":60,"accuracy":0.9836065573770492,"ingestion_time_ms":13.9,"ingested_docs":14706,"avg_retrieve_time_ms":80416.4,"avg_context_tokens":7.0,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"claude-code","model":"claude-sonnet-5","tasks":61,"solved":60,"interventions":50,"cost_usd":26.33,"turns":971,"wall_s":4905.0,"tokens_in":34885807,"tokens_out":378763},{"path":"outputs/sdebench/vanilla-claude-3/coding/boltons.json","dataset":"sdebench","run_name":"vanilla-claude-3","memory":"vanilla","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":12.8,"ingested_docs":14706,"avg_retrieve_time_ms":85563.9,"avg_context_tokens":7.0,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"claude-code","model":"claude-sonnet-5","tasks":61,"solved":61,"interventions":49,"cost_usd":27.75,"turns":982,"wall_s":5219.0,"tokens_in":35801538,"tokens_out":403470},{"path":"outputs/sdebench/vanilla-codex-1/coding/boltons.json","dataset":"sdebench","run_name":"vanilla-codex-1","memory":"vanilla","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":16.8,"ingested_docs":14706,"avg_retrieve_time_ms":51154.1,"avg_context_tokens":7.0,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"codex","model":"gpt-5.4-mini","tasks":61,"solved":61,"interventions":80,"cost_usd":32.86,"turns":1781,"wall_s":3120.0,"tokens_in":64473703,"tokens_out":1113034},{"path":"outputs/sdebench/vanilla-codex-2/coding/boltons.json","dataset":"sdebench","run_name":"vanilla-codex-2","memory":"vanilla","mode":"coding","split":"boltons","total_queries":61,"correct":60,"accuracy":0.9836065573770492,"ingestion_time_ms":18.2,"ingested_docs":14706,"avg_retrieve_time_ms":53754.1,"avg_context_tokens":7.0,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"codex","model":"gpt-5.4-mini","tasks":61,"solved":60,"interventions":83,"cost_usd":37.65,"turns":1867,"wall_s":3279.0,"tokens_in":74249355,"tokens_out":1282729},{"path":"outputs/sdebench/vanilla-codex-3/coding/boltons.json","dataset":"sdebench","run_name":"vanilla-codex-3","memory":"vanilla","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":14.4,"ingested_docs":14706,"avg_retrieve_time_ms":55793.4,"avg_context_tokens":7.0,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"codex","model":"gpt-5.4-mini","tasks":61,"solved":61,"interventions":83,"cost_usd":35.12,"turns":1869,"wall_s":3403.0,"tokens_in":68638315,"tokens_out":1239311},{"path":"outputs/sdebench/vanilla-opencode-1/coding/boltons.json","dataset":"sdebench","run_name":"vanilla-opencode-1","memory":"vanilla","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":29.3,"ingested_docs":14706,"avg_retrieve_time_ms":185123.0,"avg_context_tokens":7.0,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"opencode","model":"google/gemini-3.5-flash","tasks":61,"solved":61,"interventions":74,"cost_usd":41.66,"turns":2369,"wall_s":11292.0,"tokens_in":65435546,"tokens_out":1121445},{"path":"outputs/sdebench/vanilla-opencode-2/coding/boltons.json","dataset":"sdebench","run_name":"vanilla-opencode-2","memory":"vanilla","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":14.8,"ingested_docs":14706,"avg_retrieve_time_ms":162785.2,"avg_context_tokens":7.0,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"opencode","model":"google/gemini-3.5-flash","tasks":61,"solved":61,"interventions":76,"cost_usd":39.52,"turns":2325,"wall_s":9930.0,"tokens_in":72344900,"tokens_out":1115870},{"path":"outputs/sdebench/vanilla-opencode-3/coding/boltons.json","dataset":"sdebench","run_name":"vanilla-opencode-3","memory":"vanilla","mode":"coding","split":"boltons","total_queries":61,"correct":61,"accuracy":1.0,"ingestion_time_ms":18.8,"ingested_docs":14706,"avg_retrieve_time_ms":174832.8,"avg_context_tokens":7.0,"category":null,"mean_precision":null,"mean_recall":null,"active_passes":null,"active_total":null,"coding":true,"agent":"opencode","model":"google/gemini-3.5-flash","tasks":61,"solved":61,"interventions":69,"cost_usd":34.9,"turns":2427,"wall_s":10665.0,"tokens_in":67479034,"tokens_out":1049086}]