{"status":"ok","schema_version":"scaledmemory.public_daily_research.v1","brief":{"status":"ok","schema_version":"scaledmemory.public_daily_brief.v1","latest_edition":{"date":"2026-11-12","timezone":"Australia/Melbourne","outcome":"complete","selection_window_start_at":"2026-11-10T20:00:00Z","cutoff_at":"2026-11-11T20:00:00Z","prepared_at":"2026-11-11T20:12:06.438897Z","next_edition_due_at":"2026-11-12T21:00:00Z"},"displayed_edition":{"date":"2026-11-12","prepared_at":"2026-11-11T20:12:06.438897Z","articles":[{"id":"sha256:48230cc9925ab996b81dd0f55f46392d3aad3d715eb17d79b569a661da79d6b8","work_id":"sha256:ca32f0ff8df5c5c9f1a0fa1b46e28bbe6db7097e87c88be476bb80abfc6b9375","source_id":"sha256:3aada8c0ebafeaf9ac0630199ace0deeacc04320f9012fddb24d96e0db046521","headline":"TRIAD finds multimodal retrieval struggles with complete, evolving requests","original_title":"TRIAD: Multimodal Retrieval across Abstraction, Constraints, and State","summary":"TRIAD evaluates 17 systems across six multimodal retrieval subsets, each containing 1,000 queries and 1,000 candidates, using multi-positive mAP. Performance generally declines with hierarchical distance and negative constraints; multi-turn retrieval scores lower in 39 of 40 comparisons with paired final intent, though candidate pools highly overlap without being identical. Evaluations are zero-shot and descriptive across public checkpoints; downstream agentic task completion was not tested.","why_read":"It separates similarity ranking from satisfying hierarchy, exclusions, preserved conditions, and revised dialogue state.","publisher":"arxiv.org","source_url":"https://arxiv.org/pdf/2611.12364","format":"research","lens":"systems_product","timing":"publication_date_unknown","added_to_library_at":"2026-11-11T10:07:42.248882Z","source_published_date":null},{"id":"sha256:c6d00d6e0c2ad9d63c7faa7d0b4746076939ce43512b3976882b0b2c032925af","work_id":"sha256:f7565ff04075439e974c221735fac7b8992a7794be62cc381420eab97e62e79b","source_id":"sha256:5e9c5710f4ebc2c84454dc01c2a07a6d7516c2f9c41350ecf13da60c9deb33fd","headline":"Yuanbao-Reranker reports gains from grounded data and preference distillation","original_title":"Yuanbao-Reranker: Document-Grounded Synthesis and Preference Distillation for Text Reranking","summary":"Yuanbao-Reranker trains 0.6B, 4B, and 8B dense rerankers with document-grounded synthesis, teacher-screened negatives, and bounded preference distillation. On identical Top-100 candidate pools retrieved by Qwen3-Embedding-0.6B, results use nDCG@10 multiplied by 100: five-panel averages were 70.96, 73.63, and 74.21. Panels covered Chinese, multilingual, English, long-document, and code retrieval; results are reported by the authors.","why_read":"The evaluation ties reported reranking gains to a fixed first-stage pool, an explicit metric, and diverse language, length, and domain panels.","publisher":"arxiv.org","source_url":"https://arxiv.org/pdf/2611.10405","format":"research","lens":"systems_product","timing":"publication_date_unknown","added_to_library_at":"2026-11-11T09:26:38.836762Z","source_published_date":null},{"id":"sha256:82ed6d08d2220c20c28369ff8ee385f997fd0eb3a377e389bef54a42aa812ab0","work_id":"sha256:1af9613598dfdf28c7c4ae257b24c3610f67939b4775e749ca43aeeaa70b62cb","source_id":"sha256:13fbac3383c4ef3c2d167fbb5c398e0ceb4e0e66a16b1c129cb76f81e96fd142","headline":"AutoRefine improves code-judge calibration but exposes workflow limits","original_title":"AutoRefine: Auto-Calibrating Code Judges from Workflow Outcomes","summary":"AutoRefine uses workflow outcomes and an abstaining teacher to rewrite code-judge prompts. On 85 held-out oracle tasks, including 13 failures, rubric-judge Scott’s pi rose from 0.36 to 0.60; replay results excluded about 5 percent of tasks for timeouts or errors. On 17 independently held-out current diffs, agreement rose from 0.16 to 0.51, while the prior judge remained better on 16 older diffs.","why_read":"Results suggest outcome-based calibration can help, but small samples, one internal agent and workflow, distribution shift, and hidden failures constrain generalization.","publisher":"arxiv.org","source_url":"https://arxiv.org/pdf/2611.09291","format":"research","lens":"work_operating_model","timing":"publication_date_unknown","added_to_library_at":"2026-11-10T21:35:16.065645Z","source_published_date":null}]},"freshness":{"delivery":"latest","channel_action":"publish","timeliness":"overdue","display_is_latest_edition":true,"overdue_after":"2026-11-12T21:15:00.000Z","notice":"Latest available edition: 2026-11-12. Today's update is delayed."},"provenance":{"publication_revision":48,"channel_published_at":"2026-11-11T20:12:06.839069Z","latest_edition_sha256":"433b94e003b108e945b5295a3f6b898c80626fa3a82a3cd63b1c35a24d967dc6","displayed_edition_sha256":"433b94e003b108e945b5295a3f6b898c80626fa3a82a3cd63b1c35a24d967dc6"},"scope":{"kind":"editorial_selection","complete_library_search":false,"source_publication_dates_may_be_unknown":true},"links":{"html_url":"https://www.scaledmemory.com/research/daily-brief","json_url":"https://www.scaledmemory.com/feeds/ai-daily-brief.json","rss_url":"https://www.scaledmemory.com/feeds/ai-daily-brief.xml","guide_url":"https://www.scaledmemory.com/llms/research.md"}},"daily_sources":{"schema_version":"aikl.website_daily_sources.v1","edition_date":"2026-11-12","timezone":"Australia/Melbourne","window_start_at":"2026-11-10T20:00:00Z","cutoff_at":"2026-11-11T20:00:00Z","items":[{"id":"40d6089d-c384-4955-8123-affe61e7d452","title":"Remote Labor Index: Measuring AI Automation of Remote Work","source_url":"https://www.remotelabor.ai/","publisher_host":"remotelabor.ai","added_to_library_at":"2026-11-11T12:29:51.027057Z"},{"id":"4eede715-f3d8-4aae-ad84-4c3b7c224939","title":"TRIAD: Multimodal Retrieval across Abstraction, Constraints, and State","source_url":"https://arxiv.org/abs/2611.12364","publisher_host":"arxiv.org","added_to_library_at":"2026-11-11T10:07:42.248882Z"},{"id":"2b405319-07d4-4f03-acdd-def3ffc9b751","title":"Calibrating Document-Local Boundaries for Contextual Chunk Retrieval","source_url":"https://arxiv.org/abs/2611.12097","publisher_host":"arxiv.org","added_to_library_at":"2026-11-11T09:52:07.654174Z"},{"id":"504ccb08-3848-43b3-a9ed-7c6875ee0691","title":"Yuanbao-Reranker: Document-Grounded Synthesis and Preference Distillation for Text Reranking","source_url":"https://arxiv.org/abs/2611.10405","publisher_host":"arxiv.org","added_to_library_at":"2026-11-11T09:26:38.836762Z"},{"id":"d297c2de-29ba-4763-88b3-359b177e7223","title":"Beyond Ranking: Are Decision Models Calibrated Relevance Scorers?","source_url":"https://arxiv.org/abs/2611.11480","publisher_host":"arxiv.org","added_to_library_at":"2026-11-11T09:23:55.972691Z"},{"id":"8376062e-89cf-4f42-9086-51f7be743f27","title":"InfraRSI: Harnessing Recursive Self-Improvement for Training Infrastructure","source_url":"https://arxiv.org/abs/2611.11558","publisher_host":"arxiv.org","added_to_library_at":"2026-11-10T21:48:59.268429Z"},{"id":"9a27759d-3103-4096-bcd8-74b8795d6d5d","title":"AutoRefine: Auto-Calibrating Code Judges from Workflow Outcomes","source_url":"https://arxiv.org/abs/2611.09291","publisher_host":"arxiv.org","added_to_library_at":"2026-11-10T21:35:16.065645Z"},{"id":"e6c0869c-3cec-48d1-ba11-532b586657a5","title":"System Optimizations for Cost-Efficient Agent Harnesses","source_url":"https://arxiv.org/abs/2611.12122","publisher_host":"arxiv.org","added_to_library_at":"2026-11-10T20:41:00.050613Z"},{"id":"901e947e-5004-4454-acc5-0bc83d86c912","title":"e-means: outperforming k-means (in one-tenth the epochs)","source_url":"https://arxiv.org/abs/2611.09501","publisher_host":"arxiv.org","added_to_library_at":"2026-11-10T20:25:20.666094Z"},{"id":"2afb9083-8395-4964-b48a-8456b138f027","title":"e_means","source_url":"https://github.com/facebookresearch/faiss/tree/main/contrib/e_means","publisher_host":"github.com","added_to_library_at":"2026-11-10T20:24:57.123752Z"},{"id":"e34dfe00-9621-4cf0-a2f7-1feea403dfee","title":"AlphaProof Code: Verifying Signal’s Security at the Machine-Code Level Swarat Chaudhuri1 , Andrew Ferraiuolo1 , Anton Kovsharov1 , Tej Chajed1 , Sam Schlesinger2 , Karn Seth2 , Jonathan Katz2 , Charles Sutton1 and Pushmeet Kohli1","source_url":"https://deepmind.google/documents/36/Formal_Verification_of_Signal_4.pdf","publisher_host":"deepmind.google","added_to_library_at":"2026-11-10T20:18:25.514867Z"},{"id":"7644181a-c49c-4177-a43e-ca427b940761","title":"LongCapHarness: Capability-Guided Auto-Harnessing for Long-Context Reasoning","source_url":"https://arxiv.org/abs/2611.10463","publisher_host":"arxiv.org","added_to_library_at":"2026-11-10T20:10:37.438873Z"}],"coverage":{"complete":true,"candidate_count":12,"eligible_count":12,"omitted_count":0,"omissions":{"unsafe_url":0,"invalid_title":0,"unsafe_metadata":0},"meaning":"All active canonical metadata records admitted in this closed selection window that meet the existing public URL/title safety policy. Source bodies and private metadata are excluded."},"source_bodies_included":false,"secrets_included":false,"content_sha256":"d091219a78ac927610b48b00e348fcc67521058f805c149b0558f2512c5654ab","source_checked_at":"2026-11-22T22:24:00.095386Z"},"freshness":{"state":"retained_or_stale","source_delivery":"latest","stale_after":"2026-11-12T21:15:00.000Z","notice":"Latest available edition: 2026-11-12. Today's update is delayed. Daily links cover 2026-11-10T20:00:00Z to 2026-11-11T20:00:00Z (7 am–7 am Australia/Melbourne). 12 public links; 0 records omitted by the public safety policy. This research bundle is overdue for a successful refresh."},"scope":{"kind":"brief_and_daily_links","complete_daily_eligible_links":true,"complete_library_search":false,"archive":false},"links":{"html_url":"https://www.scaledmemory.com/accelerators/catalogue/ai-knowledge-library","json_url":"https://www.scaledmemory.com/feeds/ai-daily-research.json","rss_url":"https://www.scaledmemory.com/feeds/ai-daily-brief.xml","guide_url":"https://www.scaledmemory.com/llms/research.md"},"provenance":{"bundle_sha256":"90ca85860f77aa14226b99c7ec38564c5fbf9cb9c0459d8312575e8cf35cd162","brief_revision":48,"daily_sources_sha256":"d091219a78ac927610b48b00e348fcc67521058f805c149b0558f2512c5654ab"}}
