Long-running work accumulates material: the original goal, later constraints, attempted routes, failed experiments and current progress. Sending everything again on each continuation repeats more and more content; keeping only a summary can make the original basis of a decision difficult to recover.
+Context delivery
An Agent needs a way to look things up, too
Follow one voice task through progressive reads, JSON Pointers, exact sharing, compaction-cycle deltas, and complete recovery.
Long-running work accumulates material: the original goal, later constraints, attempted routes, failed experiments and current progress. Sending everything again on each continuation repeats more and more content; keeping only a summary can make the original basis of a decision difficult to recover.
LoopX's context delivery separates these jobs: preserve complete sources, read the facts needed for the current decision, expand supporting material around questions, and reuse unchanged structures exactly within a valid compaction cycle.
This article describes the native Codex path in released LoopX 2.2.0. Supported POSIX Codex App/SSH sessions and new benchmark heartbeat trials default to context delivery; saved off remains effective and explicit off is available. When enabled, complete content is saved first and the Agent receives a reading entrypoint for current facts. See the 2.2.0 release notes for publication and validation scope. The synthetic measurements below retain their original source and experimental conditions.
Think of a capture as a complete book. An index locates its chapters; a JSON Pointer addresses a place inside it. Progressive reading determines how deeply to open the material now. Differential delivery determines what can be reused on the next continuation.
@@ -266,6 +266,96 @@Multi-task and layered ablation: benefit chain and bo
This dual-track audit shows why control-packet byte reduction (−74.25% / −38.54%) must not be equated with whole-episode model token savings (−12.07% uncached, −7.46% total), nor used to hide early-window indexing overhead (4.05× in the first 2 hours) or cross-task counterexamples. Context Delivery’s practical value lies in slowing context bloat across 10–18 hour runs so that, together with Stride, preserved attention converts into effective late-stage experiments (see EdgeBench late-stage breakthroughs and five-arm attribution).
+Multi-task control-plane byte ratios across 8 latest Four-Arm Context ON 18h trajectories (Progressive vs. Delta)
+To separate the independent byte reductions of progressive projection (required-context.json) and in-cycle differential delivery (delivery.json) across real long-horizon runs, we audited all 8 Context ON 18h trajectories from the latest Four-Arm core suite (N = 213 model-facing wakes: 101 new-session/post-compaction baseline reads and 112 in-cycle Delta continuations):
8-task Four-Arm Context ON byte ratios (N = 213 wakes, 112 in-cycle Deltas)
| Task (Four-Arm Context ON 18h) | +Wakes (base / Δ) | +Full packets | +Progressive projection (share) | +In-cycle Delta payload (vs same-wake Full) | +Prog+Delta first-screen (body / CLI) | +
|---|---|---|---|---|---|
Grid Robot (on_stride) |
+29 (15 / 14) | +4,405,651 B | +1,166,467 B (26.5%, −73.5%) | +474,016 B (20.7%, −79.3%) | +41,878 / 76,650 B (1.8% / 3.3%) | +
VLIW Scheduling (on_stride) |
+14 (5 / 9) | +1,297,500 B | +505,826 B (39.0%, −61.0%) | +382,664 B (45.2%, −54.8%) | +30,860 / 50,152 B (3.6% / 5.9%) | +
FFmpeg Encoding (on_stride) |
+26 (12 / 14) | +2,924,522 B | +874,104 B (29.9%, −70.1%) | +759,787 B (47.1%, −52.9%) | +43,807 / 75,305 B (2.7% / 4.7%) | +
OpenTTD Transport (on_stride) |
+31 (19 / 12) | +3,288,042 B | +1,050,623 B (32.0%, −68.0%) | +436,053 B (30.8%, −69.2%) | +36,753 / 64,619 B (2.6% / 4.6%) | +
Ad Placement (on_stride) |
+29 (12 / 17) | +3,115,820 B | +1,019,704 B (32.7%, −67.3%) | +708,828 B (39.4%, −60.6%) | +64,174 / 102,437 B (3.6% / 5.7%) | +
Borden Inversion (on_stride) |
+28 (11 / 17) | +2,855,664 B | +1,001,254 B (35.1%, −64.9%) | +625,709 B (34.4%, −65.6%) | +50,602 / 90,024 B (2.8% / 4.9%) | +
Rust Workspace (on_stride) |
+15 (4 / 11) | +1,173,832 B | +439,814 B (37.5%, −62.5%) | +363,310 B (42.1%, −57.9%) | +40,956 / 64,251 B (4.7% / 7.4%) | +
ANN Vector Search (on_stride) |
+41 (23 / 18) | +3,347,659 B | +1,228,159 B (36.7%, −63.3%) | +737,968 B (38.6%, −61.4%) | +65,419 / 107,112 B (3.4% / 5.6%) | +
| 8-Run Weighted Total | +213 (101 / 112) | +22,408,690 B (100%) | +7,479,469 B (33.4%, −66.6%) | +4,488,335 B (35.8%, −64.2%) | +374,449 / 630,550 B (3.0% / 5.0%) | +
Two orthogonal byte-reduction stages appear consistently across all 8 tasks: (1) Progressive projection (required-context.json) reduces full decision packets across all 213 wakes from 22.41 MB to 7.48 MB (33.4% mean share, −66.6%, per-task range 26.5%–39.0%) by folding unselected candidates and historical long-tail detail behind on-demand JSON Pointers; (2) in-cycle Delta delivery (delivery.json) reduces the 112 in-cycle wakes from 12.53 MB to 4.49 MB (35.8% mean share, −64.2%, per-task range 20.7%–47.1%); and (3) their combined --initial-read first-screen entrypoint carries only 3.0% of raw bytes (5.0% including the CLI JSON reader envelope, i.e., −97.0% / −95.0%). In addition, the recently landed C1 unified controller/executor projection (#8266 / #8304) further reduces steady-state per-turn control-plane material by 26.7% (69.9 KB → 51.3 KB) and cuts post-writeback reprojection latency from 57.1ms to 3.3ms.
2.2.0 ships complete sources, file entrypoints, batch selection, index depth, applicable acceptance layering, exact sharing, cycle deltas and recovery. Broader decision-stage candidate, preference and Explore delivery and consistency across all entrypoints remain unfinished. Disable in Settings → This machine → Capabilities → Differential context → Codex; saved off survives upgrade. A command or new benchmark trial can use --context-delivery off. Trials do not inherit the machine switch, and existing frozen trials are not rewritten. Under the default policy, ordinary CLI, unsupported hosts and non-heartbeat benchmarks retain their full paths; an explicit unsupported on request is rejected. Unavailable hooks use full delivery. Saving configuration, actual installation and model adoption remain separate checks.
Read the same material yourself
example.py generates this complete synthetic voice task and invokes the existing reader, progressive projection and delta owners. It checks selected values, directory addresses, all ten requirements, shared references and complete recovery. It registers no Goal, admits or settles no Turn, and calls no model or speech service.
diff --git a/apps/presentation/site/public/blog/edgebench-feedback-and-memory/index.html b/apps/presentation/site/public/blog/edgebench-feedback-and-memory/index.html index 50c34077b9..c4b5d40aca 100644 --- a/apps/presentation/site/public/blog/edgebench-feedback-and-memory/index.html +++ b/apps/presentation/site/public/blog/edgebench-feedback-and-memory/index.html @@ -16,12 +16,12 @@Ruiteng Huang · October 29, 2026 · Updated November 3
What does an agent achieve with ten more hours?
How EdgeBench reshaped the LoopX control plane
A research run can start ahead, then barely improve for ten hours. It can keep changing code and lose its best result. These failures made us revisit goals, feedback, experiment size and recovery.
The work began with trace diagnosis in early October: repair evaluation and continuation, follow evidence into the next experiment, then charge the control plane for its own cost. This account follows that progression, including results that challenged the approach.
This article asks how mechanisms changed real research and where they did not help. For what actually reaches the next decision, begin with progressive and differential context delivery, then return to the outcomes and complete costs below.
Ruiteng Huang · October 29, 2026 · Updated November 9
What does an agent achieve with ten more hours?
How EdgeBench reshaped the LoopX control plane
A research run can start ahead, then barely improve for ten hours. It can keep changing code and lose its best result. These failures made us revisit goals, feedback, experiment size and recovery.
The work began with trace diagnosis in early October: repair evaluation and continuation, follow evidence into the next experiment, then charge the control plane for its own cost. This account follows that progression, including results that challenged the approach.
This article asks how mechanisms changed real research and where they did not help. For what actually reaches the next decision, begin with progressive and differential context delivery, then return to the outcomes and complete costs below.
Here, kernel means LoopX’s shared goal, admission, transaction and recovery rules. LoopX keeps the original goal, acceptance conditions, work ownership and evidence durable across sessions and hosts. Its control plane constrains state transitions and supplies feedback for the model’s next choice. It does not certify that the choice is insightful. A recorded finding, a read receipt and a replan entry are intermediate evidence; the relevant chain ends in actual work and a validated outcome.
-The TypeScript refactor article explains ownership and model-assisted semantic repair. Below, we follow how these rules affect actual research work.
-- GoalAcceptance and scope
- EvidenceResults and limitations
- DecisionContinue or change course
- WorkExecute and validate
- RetentionKeep the best valid result
We separate three observations: an initial advantage, an improvement obtained later, and retention of the best validated artifact. A higher final score can come almost entirely from a stronger starting point. A better intermediate candidate can disappear from the terminal workspace. More runtime, Todos or graph nodes establish none of these outcomes.
-1. The research progression
| Period | Focus | Change | Implementation |
|---|---|---|---|
| Early October | Feedback and starting conditions | Repair blind-run prompts, ledger and recovery defects; establish original-task, scoring and continuation boundaries. | #5586, #5844, #5879 |
| October 9–11 | Original goals and useful feedback | Preserve full evaluation responses, candidate identity and original requirements; implementation or submission is not Goal acceptance. | #6031, #6039, #6133, #6199 |
| October 12–16 | Experiment size and recovery points | Replace mechanical shrinking with grow / switch / verify; jointly preserve conversation, candidates, evaluation opportunities and budget; switch context before replanning. | #6298, #6321, #6305, #6622, #6649 |
| Mid-to-late October | Later progress and best retention | Separate starting advantage, subsequent gain and retained best in common windows; Schema exposes high cost and stalled progress. | #6440, #6532, #6671 |
| October 22–28 | Complete control-plane cost | Reduce passive polling and repeated reads; inspect original-Turn closeout, full context, invalid-baseline fallback and recovery cost. | #6983, #7013, #7036, #7197, #7269 |
| Late October–early November | Late-stage plateau breakthrough and five-arm attribution | Break the four-arm 91.94 plateau on Astra max 18h (+9 repaired declarations, 0 regressions); separate progressive deltas, Stride sizing and micro-tuning waste across ANN 18h and five-arm studies. | #7311, #7759, #7800, #7959, #7960 |
Cases are selected for mechanism changes and intelligible failures. PRs anchor phases rather than form an exhaustive daily release log. Dates describe research phases; linked records retain creation, merge and historical qualification details.
2. First make the feedback trustworthy
+The TypeScript refactor article explains ownership and model-assisted semantic repair. In implementation, every governed long-horizon iteration executes a five-stage control and feedback loop:
+- Admission & Compile
should-runboundary - Progressive / Delta ReadOn-demand current facts
- Work & YardstickBounded execution & proxy
- Writeback & Settle
EffectTurnatomic commit - Closed-Loop ReplanReplan / Stride / retain best
We separate three observations: an initial advantage, an improvement obtained later, and retention of the best validated artifact. A higher final score can come almost entirely from a stronger starting point; a better intermediate candidate can disappear from the terminal workspace. We also separate our evaluation into two progressive tiers: Layer 1 (Test-Time Compute Scaling, established)—compared with single-shot sessions or early-stopping Native Goal runs (exit=0 within 1–2h), the governed closed loop sustains 18-hour exploration and achieves decisive late-stage gains across the latest Four-Arm core tasks (ANN 7,159 vs 4,372 QPS, Grid 1,977 vs 2,430, VLIW 1,042 vs 1,159, Rust 27.18 vs 26.93, FFmpeg 2.7086× vs 2.6904×, Borden 80.07 vs Official 61.2); and Layer 2 (Iso-Budget Pareto Optimality, active frontier)—outperforming the zero-protocol-tax Official baseline (stop-hook infinite continuation + hourly native evaluator feedback naturally anchoring attention) at equal time and token budgets by satisfying the net-benefit inequality $\Delta\text{Effectiveness}(s_t, H_\theta) > \Delta\text{Efficiency Loss}(s_t, H_\theta)$.
1. The research progression
| Period | Focus | Change | Implementation |
|---|---|---|---|
| Early October | Feedback and starting conditions | Repair blind-run prompts, ledger and recovery defects; establish original-task, scoring and continuation boundaries. | #5586, #5844, #5879 |
| October 9–11 | Original goals and useful feedback | Preserve full evaluation responses, candidate identity and original requirements; implementation or submission is not Goal acceptance. | #6031, #6039, #6133, #6199 |
| October 12–16 | Experiment size and recovery points | Replace mechanical shrinking with grow / switch / verify; jointly preserve conversation, candidates, evaluation opportunities and budget; switch context before replanning. | #6298, #6321, #6305, #6622, #6649 |
| Mid-to-late October | Later progress and best retention | Separate starting advantage, subsequent gain and retained best in common windows; Schema exposes high cost and stalled progress. | #6440, #6532, #6671 |
| October 22–28 | Complete control-plane cost | Reduce passive polling and repeated reads; inspect original-Turn closeout, full context, invalid-baseline fallback and recovery cost. | #6983, #7013, #7036, #7197, #7269 |
| Late October–early November | Late-stage plateau breakthrough and five-arm attribution | Break the four-arm 91.94 plateau on Astra max 18h (+9 repaired declarations, 0 regressions); separate progressive deltas, Stride sizing and micro-tuning waste across ANN 18h and five-arm studies. | #7311, #7759, #7800, #7959, #7960 |
| Early November | 8-task byte ratios, Stride/Yardstick audit, and controller decoupling | Audit 8-task Four-Arm Context ON progressive/delta byte shares (33.4% / 35.8%); quantify Stride v2 transitions & Yardstick sibling rank reversals and fix paired candidate comparison; decouple Control/Worker sessions. | #8101, #8266, #8304, #8380 |
Cases are selected for mechanism changes and intelligible failures. PRs anchor phases rather than form an exhaustive daily release log. Dates describe research phases; linked records retain creation, merge and historical qualification details.
2. First make the feedback trustworthy
The first Portfolio experiments exposed measurement and execution defects before they could tell us much about strategy. A blind wrapper had removed iterative optimization guidance along with the judge entrypoint. A ledger check attributed earlier market returns to holdings bought only at the closing price. Both confounded the intended comparison; repairing them required new attempts while preserving the old results.
One small oracle made the accounting error clear. Start with 100,000 cash, buy 100 units at a closing price of 200, and pay 10 commission. Immediate equity is 99,990 and NAV is 0.9999; the new holding did not earn the preceding rise from 100. Independent reconciliation then matched all 243 days in each of two frozen samples, with equity residual below 4×10⁻⁹ and no share-count difference. This established ledger consistency, not investment skill. The versioned repair is documented in EdgeBench #14.
Continuation had a related failure: an outer process could wake repeatedly after a work item ended without recovering a useful next action. The governing question must return to the original goal. Completing a Todo can justify another experiment, waiting, or completion; it cannot by itself settle goal acceptance. Correct recovery and feedback make subsequent research interpretable.
@@ -122,6 +122,38 @@Early-November update: ANN 18h layered ablation and c
- Context Delivery × Stride multi-task and layered ablation (
ANN 18h, gpt-6.1-sol / xhigh, Recall@10 ≥ 0.95): Across 18-hour vector-search runs, Explore · Context on + Stride reached 7,159 QPS (+24.4% over Official at 5,753 QPS, and ahead of Explore · Context on at 6,020 QPS and Context off at 5,766 QPS). Paired control-packet selected bytes fell 38.54% (3.09 MB → 1.90 MB, with 24 delta events reduced by 74.25%), while full-18h main-session uncached input tokens fell 12.07% (8.57M → 7.53M) and total input tokens fell 7.46% (272.03M → 251.74M). Alongside these gains, we retain three explicit boundary counterexamples: project+read process stdout was 1.57×–1.83× raw selected file bytes, first-2h early-window uncached input tokens rose 4.05× (0.25M → 1.00M) during initial baseline establishment, and the cross-task Historical Grid counterexample ordered Official 1967 < Off 2054 < On+Stride 2088 < On 2242. Full charts and dual-track accounting appear in the context-delivery ablation section.
- Cross-task five-arm trajectory attribution (Portfolio, Grid, Graph, Ad Placement): In these reproduced five-arm attribution report (#7960) and cross-task inflection & replan overhead study (#7959), the five-arm taxonomy labels 86 classified replan events, while the three managed-arm totals are 44 + 36 + 24 = 104; valid-jump counts likewise differ (27 taxonomy vs 5 + 11 + 15 = 31). We therefore present 5/44 = 11.4%, 11/36 = 30.6% and 15/24 = 62.5% as arm-local ratios, without claiming one reconciled common denominator or cross-arm causal total. These are descriptive, arm-local observations rather than a matched repeated C2 comparison and do not justify a production default change. The reports also describe native-feedback retention on Portfolio (Official Native 53.87, Explore 31.19, Stride 35.82) and an observed Graph increase from 0.7194 at 2.1h to 0.7240 at 9.6h, followed by 7 post-peak attempts consuming 159.87M input tokens. The report’s fixed-classification formula evaluates to 0.74329, not its recorded 0.73903; it is a conditional estimate for that fixed component vector, not an upper bound on every linear model or proof that later work was necessarily wasted. The same reports record protocol tax from 23.9% to 12.4%–14.1% descriptively.
-
ANN 18h, gpt-6.1-sol / xhigh, Recall@10 ≥ 0.95): Across 18-hour vector-search runs, Explore · Context on + Stride reached 7,159 QPS (+24.4% over Official at 5,753 QPS, and ahead of Explore · Context on at 6,020 QPS and Context off at 5,766 QPS). Paired control-packet selected bytes fell 38.54% (3.09 MB → 1.90 MB, with 24 delta events reduced by 74.25%), while full-18h main-session uncached input tokens fell 12.07% (8.57M → 7.53M) and total input tokens fell 7.46% (272.03M → 251.74M). Alongside these gains, we retain three explicit boundary counterexamples: project+read process stdout was 1.57×–1.83× raw selected file bytes, first-2h early-window uncached input tokens rose 4.05× (0.25M → 1.00M) during initial baseline establishment, and the cross-task Historical Grid counterexample ordered Official 1967 < Off 2054 < On+Stride 2088 < On 2242. Full charts and dual-track accounting appear in the context-delivery ablation section.9. Closing the feedback loop: Stride v2 transitions, Yardstick rank-reversal counterexamples, and meta-reasoning
+Replan, Stride Policy and Yardstick are not administrative forms for the Agent to fill out; their core purpose is to establish an evidence-driven closed feedback loop—using local and native feedback to calibrate experiment step size, filter false-positive candidates, and trigger cross-family switches when an algorithm family hits diminishing returns. We audited the shipped Stride v1 → v2 transition records, the 8 latest Four-Arm Context ON trajectories, and the Yardstick rank-preservation logs:
Stride v1 (89.2% shrink collapse), Stride v2 (grow / verify / switch, N=249 transitions), and the latest Four-Arm Context ON runs (8 tasks, N=25 unique reviews: 70.6% improved, 29.4% flat guard, 0% regressed). Right (B): Yardstick trajectory audit exposing time-trend confounding (95.2% vs. 100% naive later-submission baseline), same-parent sibling rank reversal (0/2, fixed in #8380), 64-case fresh gate protection, and TLE cliff self-calibration. Open image for full size.9.1 Stride Policy: from v1’s 89.2% shrink collapse to v2’s action-to-gain loop
+-
+
- Why
Stride v1(grow / hold / shrink) was retired: Across early Grid and Borden traces (N = 37records), ambiguous relative-size labels caused the model to chooseshrinkin89.2%(33/37) of reviews whenever a local tweak failed to immediately beat the incumbent best (grow10.8%,hold0%), trapping the next 9 hours in single-constant micro-tuning.
+ Stride v2(verify / switch / grow, #6321 / #6532 / #6654) distribution and transition matrix (N = 372records,249paired transitions): +-
+
grow(gated onbasis = supported_progress):79.4%next-step score improvement (27/34 improved, 6 flat, 0 regressed); e.g., onAd Placement, three consecutivegrowsteps on composite annealing and warm restarts yielded three consecutive native gains (up to+954,331in a single step).
+ verify(discriminating experiment + baseline guard):58.8%blocked noisy local false-positives atflat(97/165 protecting Incumbent Best), while40.0%converted into score gains (66/165 improved) and only0.6%regressed.
+ switch(cross-family algorithm switch): cross-family exploration carries high variance (74.0%protected atflat), yet the16.0%(8/50) hits delivered decisive step-change escapes after prior families plateaued (e.g.,Graphswitching from 2-hop SGGC to PPR multi-scale diffusion andC&Sresidual propagation:0.7194 → 0.7225 → 0.7240).
+
+- Latest Four-Arm
Context ON8-task audit and Good Cases (N = 25uniqueStride v2reviews,17scored transitions): With progressive/differential context delivery enabled (33.4% / 35.8% byte share across 8 tasks), theon_strideruns recordedverify 16 (64.0%) / switch 6 (24.0%) / grow 3 (12.0%), with76.0%(19/25) explicitly couplingExploreledger evidence,100%(17/17) full execution, and70.6%next-step improvement (12/17 improved),29.4%flat guard (5/17 flat), and0%regressions: +-
+
OpenTTD($11.28M → $34.59M, 3× gain):grow(double-track trunk expansion$11.28M → $16.96M) →verify(signal spacing and fleet sizing$16.96M → $28.14M → $30.71M) →switch(full-map long-haul coal/iron network, jumping to$34.59M).
+ Borden(7.01 → 80.07, ahead ofOfficial 61.2): Twoverifysteps advanced multi-start L-BFGS inversion from7.01to77.96; when review #3 hit a local plateau (77.96 → 77.97),switchpivoted to coordinate descent with cross-well sensitivity reweighting to reach80.07.
+ FFmpeg(0.90× → 2.7086×, overall best):switchinto x264 preset and thread scheduling (0.90× → 2.52×) →growalong thesubme/me/trellisrate-distortion frontier (2.52× → 2.66×) →verifycalibrating CRF and GOP boundaries to2.7086×(ahead ofOfficial 2.6815×andNative Goal 2.6904×).
+ Grid(2,288 → 1,977, lower is better) &VLIW(1,266 → 1,042, lower is better):Gridusedswitchto pivot into hybrid cost tables and C++ beam-search repair, dropping steps from2,288to1,977(453 steps fewer thanNative Goal 2,430);VLIWusedverify → growto move from greedy list scheduling (1,266) to modulo-scheduled loop pipelining (1,042vs.Native Goal 1,159).
+
+
9.2 Yardstick trajectory audit: three counterexamples and paired-candidate repair (#8101 / #8380)
+Under the 1-hour native submission cooldown (native3600), Yardstick maintains a work-item local evaluation ruler and reconciles it against official Native outcomes (#7037 / #7651 / #8101). Auditing real trajectories surfaced three counterintuitive statistical findings and directly motivated a control-plane repair:
-
+
- Apparent global agreement confounded by time trend vs. same-parent sibling reversal (#8101): On the Grid panel, global pairwise agreement between local Proxy and official Native looked as high as
20/21(95.2%) and10/10(100%), but this was heavily confounded by the temporal trend of improving code—a naivelater_submissionguess achieved21/21(100%). Once time trend is stripped out by inspecting same-parent sibling candidates forked fromv10(native1958) inOfficial Grid v11 / v12, the local 20-seed mean predicted improvement for both (-2.95and-1.40steps) while official Native regressed on both (1958 → 1966 [+8]and1962 [+4], same-parent rank agreement0/2).
+ - Large-sample local significance still reverses vs. fresh gate protection (
Ad v30vs.Ad r11):Ad v30beatv28across 64 local seed pairs (+675K), 16 holdout pairs, and the 80-pair combined set (paired $t = +2.41$), yet regressed by-60,072on the official 50-case evaluator; conversely, inAd r11, a 64-case fresh gate blocked negative candidatev28(-77,049) and protectedv22at77.63(+27.22Mabove the initial submission). Local yardsticks therefore act as high-recall negative gatekeepers and same-parent tie-breakers, not substitutes for native feedback on small mean deltas.
+ - Self-calibration against TLE timeout cliffs via Yardstick: In
Grid Stride, the 2nd submission improved the local 8-seed mean (345 → 339) but scored0natively due to a single case timing out at30.085s; reading back the counterexample, the model upgraded its local Yardstick to “60 normal runs (24s cap) + 60 stress runs (T=X budget gate)”, immediately restoring a100%pass rate and reaching2,088native steps (Ad v45 → v48similarly tightened its local budget ruler from4.62sto4.42sto eliminate a1/50case5.107stimeout).
+
Architectural repair for same-parent reversals and unsubmitted probes (#8380): Previously, measuredSearchComparison and projectExploreMeasurementFeedback compared the last two local proxy records against the last two official goal records by array position. Whenever a rejected sibling proxy false-positive (v11) or a subsequent local-only screening probe (v13) sat in between, positional comparison mismatched candidate identities and suppressed proxy_order_reversed: true. #8380 anchors local proxy comparison to the exact [goal.previous_best, goal.latest] candidate pair so sibling rank reversals reliably trigger yardstick recalibration.
9.3 Next horizon: decoupling Control Session from Worker Session and meta-reasoning
+Moving from Layer-1 compute scaling to Layer-2 iso-budget Pareto optimality points to two structural upgrades: first, decoupling the Control Session from the Worker Session (C1–C6, #8266 / #8304) so Replan, Stride selection and receipt writeback execute in an isolated control session while the worker keeps a zero-pollution KV cache and continuous domain focus; second, evolving the control plane into a Value-of-Computation (VoC) meta-reasoning core (Selecting Computations) that combines risk-sensitive yardsticks (worst-case tail-latency margin + paired rank calibration) with cross-family switch triggers so every unit of control overhead buys a better next experiment.