Add estimate-marking awareness and a second fact-check round (E3) #130

Merged
weichen merged 1 commits from feature/agent-e3-evidence-gate into main 2026-10-10 16:10:42 +00:00
Owner

E3(提案 §2 D4):①标注感知豁免——分句判定「目标/预计/假设」邻域内的数字放行(固定窗口会误放行相邻句数字,被既有测试抓住);②修正轮升两轮——第二轮仅在未核实集合有变化时触发(共享续写预算不被重复拦截挤占,同样被既有测试约束);③base.deliverables v6 数字溯源硬规则 + research/explore/general 子代理引用格式;④eval 报告新增 unverifiedUnmarkedFacts(与后端同一标注口径)。

Evidence: go test ./... 全绿 + 3 个新回归锚点。行为面(提示词+门禁)eval 前后对照在 Dev 侧做(裁定 6)。

Refs: Dev 任务 feature/agent-e3-evidence-gate。

E3(提案 §2 D4):①标注感知豁免——分句判定「目标/预计/假设」邻域内的数字放行(固定窗口会误放行相邻句数字,被既有测试抓住);②修正轮升两轮——第二轮仅在未核实集合有变化时触发(共享续写预算不被重复拦截挤占,同样被既有测试约束);③base.deliverables v6 数字溯源硬规则 + research/explore/general 子代理引用格式;④eval 报告新增 unverifiedUnmarkedFacts(与后端同一标注口径)。 Evidence: go test ./... 全绿 + 3 个新回归锚点。行为面(提示词+门禁)eval 前后对照在 Dev 侧做(裁定 6)。 Refs: Dev 任务 feature/agent-e3-evidence-gate。
weichen added 1 commit 2026-10-10 16:08:36 +00:00
Add estimate-marking awareness and a second fact-check round (E3)
ci / vet-and-test (pull_request) Successful in 41s
06f5835211
Why: The E1 live comparison was the only metric that got worse
(unverifiedFacts 7 -> 12): the gate gave one correction round and then
shipped bare numbers. The design (E3) calls for number sourcing with a
mechanically checkable marking escape: numbers labeled as
targets/assumptions are honest planning quantities, not fabricated
facts, so the gate must let them pass - and the model must be told that
marking is the sanctioned way out.

Why: The marking decision is clause-scoped, not a fixed window - a
window wrongly blessed "3,500" because a neighboring clause said
"target 85%" (caught by the existing test). The second round only
fires when the unverified set actually changed, because the
continuation budget is shared across all final gates and a repeat
block on identical tokens starves the downstream gates (also caught by
an existing test).

Evidence: go test ./... green; new anchors TestMarkedAsEstimate,
TestUnverifiedFactsMarksPass, TestFactCheckGateRounds; the live eval
gains an unverifiedUnmarkedFacts field sharing the same marking rule.

Impact: base.deliverables v6 adds the hard number-sourcing rule;
research/explore/general sub-agent profiles v+1 require citations for
key numbers; fact gate: marked-in-clause numbers pass, round 2
escalates to "last interception" with the marking instruction, then
passes through with the check line. Eval report keeps the strict
series and adds the E3-series field.

Refs: proposal 2026-10-10-agent-delegation-routing-sandbox.md (E3,
D4); Dev task feature/agent-e3-evidence-gate.
weichen merged commit bd687d27d4 into main 2026-10-10 16:10:42 +00:00
Sign in to join this conversation.
No Reviewers
No labels
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: weichen/Agent-System-Backend#130