Research problem entry

Everything starts with a research question.

这个页面是 10 Paper Portfolio 的研究问题入口。它不是论文标题清单，也不是 30 篇文章的承诺表；它是一个 v0 paper sprint harness 的问题层： 每个槽位先回答“什么问题值得问”，再进入 evidence、SOTA pressure、strongest objection 和 kill / merge rule。

30research-question slots
4today priority slots
RQstable question ids
Ppaper handles may drift

今日优先级

当前不是平均推进 30 个槽位，而是用少量高压问题带动整个 board。每个 priority slot 都必须有 48h evidence path。

RQ01review wedge

观测数据应如何进入因果建模？

当 heterogeneity is primitive 时，因果建模不能从平均效应再补异质性开始。 下一步是 `outline.md` 和 30 篇 source matrix。

RQ03sprint

Causal mind 是否真的改变工作决策？

P02 已有内部 10-event table；下一步要做 independent recoding 或更少自指的样本。

RQ05sprint

Agent runtime probe 需要什么 failure semantics？

P04 有 3 行 replay 证据；下一步是 broader case 和 verified opponent metadata。

RQ09harness

什么状态机能防止论文候选无限漂浮？

P08 已改变 merge / park 决策；下一步要用整个 RQ board 作为 state-machine input。

问题进入规则

新问题进入 board 前，先回答下面五个问题。答不上来，就保持 open slot，不要急着写论文。

它改变什么理论、决策、系统能力或投稿策略？
48h 内什么证据能让它不只是好听？
最强 paper-killing objection 是什么？
它的 source 属于 Primer、HCGM、DiscoSCM、Team OS、old draft 还是 newly scouted？
需要 gong 深介入，还是 agent 可以先跑 evidence pass？

30 Research-Question Slots

`RQxx` 是稳定问题槽位；`Pxx` / old `A2` / `B5` 只是候选或论文 handle。 A/B/C package role 必须等 evidence 成熟后再浮现。

Slot Research question Handle State 48h evidence Kill / merge rule
RQ01When should causal modeling start from observational data, and how does the answer change if heterogeneity is primitive?P01 observational surveyreview wedgeOutline plus source matrix with 30 core references.Kill expansion if it becomes a generic causal inference method list.
RQ02Can structural causal mechanism interfaces improve LLM / agent counterfactual reasoning beyond prompt-only baselines?old P01 / A1liveOne benchmark task where mechanism structure changes answer or failure attribution.Park if novelty is only prompt engineering around existing benchmarks.
RQ03Can agent causal reasoning be evaluated by decision deltas in real workflows?P02 / old A2sprintIndependent recoding or broader sample for the 10-event decision-delta table.Downgrade to internal eval if deltas are post-hoc or founder-specific.
RQ04Can collaboration traces become causal data for improving multi-agent workflows?P03 / old A3liveRedaction-safe event schema over 20 events and one intervention contrast.Park if privacy redaction destroys signal or examples stay anecdotal.
RQ05What failure-class semantics do agent runtime health probes need to avoid wrong escalation?P04 / old B1sprintOne less self-referential replay case plus verified opponent metadata.Downgrade to runbook/blog if taxonomy is standard SRE with no agent-runtime contribution.
RQ06Do seed-to-action trust gates reduce fake progress in agent-managed work?P05 / old B2componentCompare 10 tasks with and without evidence/trust gates.Keep merged into RQ09 unless independent evaluation signal appears.
RQ07Can sparse human taste be converted into durable agent work changes?P06 / old B3parkedFive real feedback events with artifact_delta or decision_delta tracking.Park until real feedback changes downstream work.
RQ08Can app-less agent harnesses optimize communication bandwidth better than dashboard-first workflows?P07 / old B4parkedObservable bandwidth rubric over three workflows.Park if the metric remains taste-only or too subjective.
RQ09What state machine helps multi-agent paper production decide sprint / park / merge / kill?P08 / old B5sprintConvert top candidates into structured entries and prove work allocation changed.Kill as paper if it adds bureaucracy without changing decisions.
RQ10Can Causality Primer become an agent skill curriculum with task-level evaluation?P09 / old C1parkedOne chapter becomes one skill exercise plus eval.Keep as skill asset unless it gains research contribution.
RQ11What do tabular foundation models actually learn, and what changes under a DiscoSCM unit/context/noise/query audit?P11 tabular foundation modelsreview wedgeRun one repeated-unit/grouped/context-shift benchmark cell after the source and manuscript wedge.Park as a reader asset if the audit does not change model classification, benchmark design, or a testable proposal.
RQ12What agent profile state is minimally necessary for team intelligence?old C3 / component of RQ04componentBefore/after routing-error comparison.Keep as Team OS background unless it creates independent evidence.
RQ13Does a kill matrix prevent AI-generated paper-shaped artifacts?P12 / old C4componentApply reviewer objections to 15 candidates and log changed decisions.Keep merged into RQ09 unless it becomes independently measurable.
RQ14How can team agent logs preserve causal signal without leaking private context?P13 / old C5parkedRedaction levels over three concrete log examples.Park until publishable examples exist.
RQ15Can probability distributions be directly aggregated by a defensible information aggregation operator?WEDGE-001liveTwo-page SOTA/opponent memo with 10 closest related works.Park if the operator reduces to known pooling/fusion or lacks advantage.
RQ16How can the first bounded DiscoSCM slice learn a token-modulated outcome mechanism from one factual observation per token?Learning DiscoSCM · P16 supporting seedflagship scienceV2 logic contract and factual-law/non-identifiability/Cauchy/off-diagonal propositions; next run is a truth-isolated alternative-query benchmark.Not an active submission package; require comparator-aware evidence and do not infer Layer 3 identification from factual-law recovery.
RQ17What is the minimal benchmark for DiscoSCM-style counterfactual / intervention reasoning?DiscoSCM evalopenThree tasks, baseline, scoring rule, failure taxonomy.Park if tasks duplicate existing causal LLM benchmarks.
RQ18Can CausalQwen / CausalLLM evidence be rebuilt into a credible causal-reasoning package?CausalLLMopenLocate current evidence, dataset, baseline, and reproduction path.Park if evidence cannot be reproduced.
RQ19When are population-level cycles projections of unit-level DAG mechanisms?Primer Ch03openFormal note plus examples separating projection cycles from equilibrium semantics.Merge into Primer if no paper-level opponent appears.
RQ20Can robust prediction under distribution shift be reframed as mechanism preservation across heterogeneous units?causal representationopenThree SOTA opponents plus one toy shift example.Park if it is a generic distribution-shift survey.
RQ21Can causal mechanism diagnosis improve LLM / agent failure attribution?Causal intelligenceopenTwo failure cases where mechanism diagnosis changes repair path.Merge into RQ02/RQ03 if not independently sharp.
RQ22How should no-data causal reasoning be represented before observational evidence enters?P01 source bucketopenSource matrix examples from action, physics, social mechanism, and counterfactuals.Merge into RQ01 unless it demands standalone treatment.
RQ23What changes when unit-specific treatment-response mechanisms become primary objects beyond HTE / ITE?P23 / HCGM causal-effectpreflightUpstream theory, robustness, IHDP and WAWS evidence plus HTE/ITE comparison.Reposition if the object-level distinction adds no value beyond direct robust or representation learners.
RQ24Can PyWhy / DoWhy-style workflows become executable teaching/evaluation assets for causal intelligence?Primer toolingopenOne model-identify-estimate-refute case connected to Primer.Keep as course/tooling asset unless evaluation becomes novel.
RQ25Can old manuscript assets be upgraded through a standard SOTA/opponent pass?old draft protocolopenInventory five old drafts and run one complete upgrade/park memo.Merge into RQ09 if it remains production protocol.
RQ26Which research problems should gong personally deep-touch, given a 2-3 slot attention budget?owner attentionopenTaste/novelty/owner-need rubric applied to all live slots.Keep internal unless it becomes RQ09 evidence.
RQ27Can a source matrix prevent survey papers from becoming literature lists?P01 processopenBuild source matrix for RQ01 and record how it changes outline.Merge into RQ09 if it is only workflow evidence.
RQ28Can a single-paper entry site improve public feedback and reviewer readiness?public projectionopenOne P01 or RQ09 entry updated from placeholder to real public-safe page.Keep as delivery asset unless feedback effect is measured.
RQ29What is the publication boundary between WeHub internal evidence and public research claims?research integrityopenBoundary memo using P02/P04/RQ14 as cases.Merge into RQ14 if it becomes only log privacy.
RQ30How should WeHub mine new research questions continuously instead of waiting for brainstorming bursts?question engineopenDaily question intake protocol plus one week of changes to board state.Keep as internal loop unless it produces publishable method evidence.