You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
T2 distillation repeatedly fails turn integrity, including singleton blocks #5283
With pi and billion-context 0.1.209 (e3a3cbd), T2 distillation repeatedly rejects every tested range with turn-integrity errors, including a single existing T1 block. No compression is applied, while automatic emergency nudges repeatedly recommend the same failing operation. Ordinary fresh raw-message folding still works when it does not overlap a T1 block.
Reproduction observed
Session contains 67 active T1 blocks (~30K summary tokens), plus substantial raw tool history.
Broad block distillation fails. Smaller b86–b91, singleton b85 and singleton b2 also fail. Header-string and JSON-encoded startId/endId forms give the same result.
Representative error: Merging b2 cannot satisfy turn integrity without unfolding already-folded message(s) owned by those blocks (...) — doing so would drop their existing coverage. Align the range to whole turns (do not cut through a reasoning/tool-call/result group) or merge fewer blocks.
Status still lists the blocks as active distillation targets. Its raw-range advice changes, but those successful raw folds do not resolve T2 eligibility.
A supported way to replace an existing block's summary without losing its original coverage, or actionable eligibility advice that can succeed. At minimum, do not repeatedly recommend unavailable T2 ranges after deterministic turn-integrity rejection. No private conversation or project details are included here.
Recovery update: a chronological boundary range b123–b137 successfully distilled all 68 active blocks (including b138, whose source was chronologically earlier). b1–b138 was NOT the full chronological interval: IDs were creation-order rather than current source order. acp_status now confirms one T2 block, ~2.3K summary tokens, and an idle nudge. The remaining defect appears to be misleading numeric-range guidance plus insufficient diagnostics for locating the true earliest/latest anchors, rather than an unconditional T2 failure. No integrity guard was bypassed.
误导性诊断(缺陷):报错里的 (r1, c1) 样例列的是被合并块自己的覆盖(unit-projection.ts:419,sample = consumedCoverage.slice(0,5)),不是真正违约的 ref;建议语"Align the range to whole turns"在此场景不可执行——违约组在范围之外,任何对齐都修不了它。
Still reproduces on billion-context 0.1.211 (e226544), Pi: 80 active T1 blocks (~19K summary tokens; 103% surface). Attempted broad b30–b327 after ending the preceding assistant turn and folding available raw ranges; rejected before folding: "cannot satisfy turn integrity without unfolding already-folded message(s) owned by those blocks (h_c1411834cd288b47, h_67749e1e0ba29f77, h_9afd31041063ac02, h_106efbeaf94d2fe4, h_ccdda38cb1902814_2)". Smaller recent range b319–b325 also failed in the same session; new raw ranges still fold. Repeated emergency nudges recommend b30–b327 after rejection. The prior workaround of chronological boundaries is not discoverable from the public acp_status overview (blocks sorted by size, no source range listing), so diagnostics/eligibility remain blocked. No private project/session contents included.
Activity
TimBriseby commentedon Oct 31, 2026
Recovery update: a chronological boundary range
b123–b137successfully distilled all 68 active blocks (including b138, whose source was chronologically earlier).b1–b138was NOT the full chronological interval: IDs were creation-order rather than current source order. acp_status now confirms one T2 block, ~2.3K summary tokens, and an idle nudge. The remaining defect appears to be misleading numeric-range guidance plus insufficient diagnostics for locating the true earliest/latest anchors, rather than an unconditional T2 failure. No integrity guard was bypassed.ranxianglei commentedon Oct 31, 2026
(ai: vllm-qwen/qwen3.8-27b)
查重: 不重复。相关条目:#5294(另一 agent 会话从本 issue 派生的决策请求,同一问题家族,已互链)。
标签:
bugpiscope/kernelp1— 受影响的会话里分层蒸馏能力整体停摆、上下文只增不减(p1 第②类:失控增长)。已验证:最小复现成立(仅用内核公开 API)
用
createCore/applyCompression构造最小场景(推理+工具调用回合在结果尚未上链时折叠成块,结果随后才到达),得到与你完全一致的报错文本。机制链分四层:projectUnit合并分支对全局折叠集(本块覆盖 ∪ 所有其他活跃块覆盖)做回合完整性扫描(kernel/src/unit-projection.ts:365)。只要任何一个活跃块带陈旧分裂,任何消费任意块的合并都会要求展开那些"已被消费块拥有的 id"——而纯块范围没有可展开的直接 id → 确定性致命错误,永远失败。一个陈旧块毒化全部 67 个合并,不是 67 个独立故障。我的复现里,只覆盖两条纯文本消息(全程无工具/无推理)的无辜块 b2,单块合并也报同样错误、且报错点名 b2 自己。(r1, c1)样例列的是被合并块自己的覆盖(unit-projection.ts:419,sample = consumedCoverage.slice(0,5)),不是真正违约的 ref;建议语"Align the range to whole turns"在此场景不可执行——违约组在范围之外,任何对齐都修不了它。pendingByTier(kernel/src/compress.ts:1754)对全部活跃 T1 块求和(~你的 30K),不做可行性筛查,紧急 T2 需求持续存在;压缩断路器触发后经过静默窗口衰减解除,需求再次出现——正对应你看到的"breaker 说停,后续紧急 nudge 又要求 T2"。方案评估与建议
#5294 的 A/B/C 选项修的是建议层(创建顺序 vs 源序跨度的分叉、建议前预筛),保留严格门禁。但我的复现表明这不够:即使建议选得完美,无辜单块也会被拒(b2 案例),所以 A/B/C 单独无法交付你期望的"在不丢失原覆盖的前提下替换已有块摘要"。
我建议 D = A + 门禁放宽(存量违约豁免):合并门禁只对本次合并新引入的违约要求展开——与合并前的全局扫描结果对比,存量违约(线上本来就在发送的状态,纯块跨度对它字节级保持不变,可证:carve 只触碰直接 id,消费覆盖原样透传,折叠集成员不变)放行;新违约仍诚实拒绝,保留 #4484 的保护(永不展开消费块拥有的 id);同时把报错改成点名真正未解决的 ref。
行为变更披露(D 若采纳):今天硬失败的合并将成功,并逐字保持既有分裂状态(kernel 测试
issue4484用例 3 从"必须失败"翻转为"成功且保持既有分裂")。状态:内核改动按边界规则需你拍板(A/B/C/D),实现等决策。证据与 D 的建议已同步评论到 #5294。
ranxianglei commentedon Oct 31, 2026
(ai: vllm-qwen/qwen3.8-27b)
补一行分账信息:主题类别 B(折叠预算与资格判定)——缺陷核心在合并资格判定(回合完整性门禁的全局耦合)+ 陈旧覆盖的产生路径;若按"压缩引擎行为"口径也可归 C,以讨论帖定义为准。最小复现脚本已留在本会话工作区(
.tmp/repro5290.mts,仅用内核公开 API),owner 拍板后可直接转成kernel/tests/issue5290-*回归钉。applifaction commentedon Nov 2, 2026
Still reproduces on billion-context 0.1.211 (e226544), Pi: 80 active T1 blocks (~19K summary tokens; 103% surface). Attempted broad b30–b327 after ending the preceding assistant turn and folding available raw ranges; rejected before folding: "cannot satisfy turn integrity without unfolding already-folded message(s) owned by those blocks (h_c1411834cd288b47, h_67749e1e0ba29f77, h_9afd31041063ac02, h_106efbeaf94d2fe4, h_ccdda38cb1902814_2)". Smaller recent range b319–b325 also failed in the same session; new raw ranges still fold. Repeated emergency nudges recommend b30–b327 after rejection. The prior workaround of chronological boundaries is not discoverable from the public acp_status overview (blocks sorted by size, no source range listing), so diagnostics/eligibility remain blocked. No private project/session contents included.
ranxianglei commentedon Nov 2, 2026
(ai: vllm-qwen/qwen3.8-27b)
谢谢补充——0.1.211 的数据点很有用。已对照当前 master(0.1.211 之后)重新验证:
kernel/src/report.ts的renderCompressedDrilldown),没有每块的源范围或时间位置;连按块 recap 里的 "range" 字段也只是消息条数。所以"按时间边界选范围"这个绕行确实无法从公开面发现,而试探式选范围每次失败还会喂断路器预算。这个诊断缺口正是决策请求 [PR] fix(storage): a live second writer no longer blocks startup — sharing a local-disk index.db is a supported topology (rev | fix(storage): 活着的第二写者不再阻止启动——共享一个本地盘 index.db 是受支持拓扑(撤回 #5213 的拒绝策略) #5294 选项 A③ 的内容。状态不变:修复集等 owner 在 #5294 拍板(A/B/C/D;D = 建议层修正 + 合并门禁对存量违约豁免 + 报错点名实际违约 ref + acp_status 暴露块源位置)。你的确认已记录到 #5294。
ranxianglei commentedon Nov 5, 2026
(ai: vllm-qwen/qwen3.8-27b)
新实例(dsh 车道,v0.1.212)—— #5746「折叠卡住」,按同症状一单原则并入本 issue。
owner 在该线程的原话诊断与我们一致:「原因不是大小……坏的是它那道『折叠边界』的检查」——换了几种切法全被门禁挡回。
新日志证据(会话 e2211607…,dsh,v0.1.209→0.1.212):
本 issue 清单项现状:pi 实例(#5283 原始)、dsh 实例(#5746)。仍等 A/B/C/D 拍板。
实现协调提醒:在途 kernel PR #5700(T1 harvest stall escalation,fixes #5698/#5671)压着
kernel/src/compress.ts;在途 kernel PR #5668(nested compression state integrity,closes #5667)压着 projection 面——将来落地 D 时这两个都是热文件。ranxianglei commentedon Nov 5, 2026
(ai: vllm-qwen/qwen3.8-27b)
收到 dsh 实例并入信息。核对后三点确认 + 一个编号更正:
1. 编号更正(防止后续串错线程):该评论里的「#5746」是上游 GitHub 编号 —— 我们的镜像里 #5746 是另一张单(ingest O(history) 存储问题,与本机制无关)。本单自身的 GitHub 编号也是 #5283(镜像号 ≠ GitHub 号)。后续引用请统一写「GitHub #N」。
2. 关键证据核对:
3. 实现协调已核实:点名的两个在途 PR 都真实存在 —— #5700(T1 harvest stall,分支
2026-10-09_nudge-tier-slot-rotation,压 compress.ts)与 #5667 相关(分支2026-11-04_nested-fold-ledger-fix,压 projection 面 + acp_status 展示面);另全量扫描远端分支还有约 20 个在途分支触碰这批热文件(compress/unit-projection/turn-integrity/recommend/boundaries)。无论拍板哪个选项,落地时都要对最新 master rebase 并与这些分支协调顺序。清单项现状:pi 实例(原始帖,GitHub #5283)+ dsh 实例(GitHub #5746),两实例机制一致,在本单内以子缺陷清单归并管理。仍等 owner 对 A/B/C/D 拍板(决策请求 #5294);上述新证据已同步到该线程。
ranxianglei commentedon Nov 6, 2026
(ai: vllm-qwen/qwen3.8-27b)
第三个生产实例(hermes 车道)补入本清单: 镜像 #5778(GitHub #5776)。
b102..b165失败("cannot preserve existing coverage and turn integrity");09:53–09:56 亦有重复的 T2 nudge + 失败合并;fresh message folds 正常;git diff 84fa9af5 origin/master六个 kernel 文件(compress/nudge-text/unit-projection/boundaries/recommend/report)零差异——本实例与前两实例(v0.1.209 pi / v0.1.211 dsh)及 master 同代码路径,机制结论直接适用;src/stream.ts:94,261),从不给校验过的块区间——与机制分析中建议面/恢复面缺陷一致。注意:决策请求线程 GitHub #5332(镜像 #5294)已于 11-02 关闭且无 owner 决策记录;按「决策不开新 issue」规则,决策请求已在镜像 #5778 重发(选项 A/B/C/D,推荐 D,附八问与行为变更披露)。本实例不改变机制判断与选项对比,只提高紧迫度。
ranxianglei commentedon Nov 8, 2026
(ai: vllm-qwen/qwen3.8-27b)
清单更新——第 4 个生产数据点(pi 车道,2026-11-08,报于镜像 #5778 / GH#5776): 失败批次的恢复回执本身把刚被拒的同一组范围原样再列为"唯一合格选项"。该批次 3 个范围(m01107–m01141 / m01146–m01150 / m01167–m01250)整体被拒:两个死于 merge gate(消费 b68,b69 与 b71-b73),中间那个是纯 raw fold、通过校验但被原子性陪跑;拒绝后状态不变,host 恢复提示重算后逐字再推荐同一列表("use these refs exactly as listed")。循环形状已在 master 代码级钉死:kernel/src/compress.ts:840(批量拒绝回执 "Fix or drop the offending range(s)")与 src/stream.ts:1381→compressionRecovery(:103-138)(重算+再推荐)指令自相矛盾;且 recommend 侧的 projectUnit 预筛(recommend.ts:659-732,#3236)放行了这两个范围 → 筛查与执行之间存在残留判决差,泄漏子机制需会话账本钉死。
按报告者要求已并入 #5778 决策请求为 facet ④(恢复回执排除刚被拒 span、保留原子性陪跑的合格 span+关闭筛查/执行判决差+offer 陈旧守卫)。选项 A/B/C/D 结构不变,推荐仍为 D,待 owner 拍板。