CoWAM is a selective intervention layer that represents synchronization, role, and collision obligations as coordination contracts. Each contract combines typed admissibility predicates, event-conditioned learned evidence, and calibrated intervention gates. The nominal action remains in effect unless an alternative satisfies every active obligation, remains low-risk, preserves task utility, and clears the calibrated intervention thresholds. Otherwise CoWAM preserves the nominal action or invokes a predefined abstention fallback.
This contract view separates proposal generation from intervention. CoWAM neither trains a new action generator nor changes the candidate pool. Every selector instead receives the same ordered candidates, commits before simulator outcomes are available, and is audited by one shared oracle-labeling pass. This protocol is referred to as outcome-blind same-pool evaluation. The evaluation therefore distinguishes coordination-valid selection, false and harmful intervention, natural closed-loop success, and proposal headroom.
World-Action Candidate Interface
At replanning time, a frozen proposer returns an ordered candidate pool. The controller returns one of three decisions: preserve executes the nominal action, override executes one verified alternative, and abstain invokes a task-defined fallback when no candidate is admissible. This makes policy intervention, rather than future generation, the object of the method.
Coordination Contracts
Inactive event types impose no constraint, whereas every active predicate must pass. Predicates use paired action and future evidence: examples include bounded inter-arm distance, compatible object assignments, synchronized contact progress, and nondivergent shared-object motion. The contract also stores calibrated decision thresholds and the fallback associated with contract failure.
Qualitative Cases
Three task-level examples illustrate the method. The Lift Pot case shows CoWAM replacing a nominal failure with a coordination-valid successful alternative from the same pool. Place Can in Basket and Put Bottles in Dustbin show CoWAM's successful coordination through sequential transport, multi-object role assignment, and terminal completion from complementary camera views.
Decision Procedure
The complete decision procedure is:
Receive and persist the ordered paired candidate pool.
Instantiate the active coordination contract from task and event state.
Evaluate typed predicates and event-conditioned evidence for every candidate without access to outcome labels.
Construct calibrated risk and utility bounds and evaluate all selective gates.
Override with the highest-scoring eligible alternative; otherwise preserve the contract-valid nominal action or abstain through the frozen fallback.
Persist the selected index, decision mode, contract outcomes, scores, bounds, and candidate-pool digest before shared oracle evaluation.
The same interface defines all ablations. Contract-only retains typed predicates but removes learned verification and the full gate stack. Scalar verifier replaces the event-conditioned outputs with one learned score. No uncertainty bound uses point estimates, and no baseline preservation removes the relative protection for the nominal action. Only Full CoWAM is the proposed method.
Outcome-Blind Pairing
Each evaluation unit materializes one ordered candidate pool shared by every selector. Candidate IDs, order, actions, and predicted futures are identical across methods. Selectors write their chosen index and complete decision record before an oracle call. One subsequent simulator batch labels every candidate. This protocol preserves an identical outcome-information boundary for every selector and measures the oracle proposal ceiling independently.
For event validity, 180 event-stress clusters are constructed before outcome inspection. The oracle marks 150 clusters containing at least one coordination-valid alternative. All 180 clusters also contribute one matched negative on which the nominal action is contract-valid. Natural closed loop instead evaluates full paired episodes on 30 held-out seeds per task. Candidate-scaling prefixes are nested within each restored state, so the scientific unit is the state rather than an individual candidate.
Metrics and Statistical Tests
Strict task success is simulator completion within the fixed horizon. Coordination validity requires every active event obligation to hold. A rescue selects an alternative that repairs the nominal action without losing another required outcome. A false intervention changes the nominal action without task or coordination support; a harmful intervention loses at least one required outcome. Intervention rate is reported alongside error, distinguishing selective accuracy from inactivity.
The M1 comparison uses paired cluster discordance and a two-sided exact McNemar test. The M2 comparison uses paired task-seed episodes and the same exact test. Candidate records within a pool are not treated as independent. Ranking metrics are computed on task-seed-disjoint groups; calibration is measured by expected calibration error and Brier score. Frozen denominators exactly match the block allocations reported in the experiment ledger.
Denominator Ledger
The ledger counts independent scientific units rather than summing every reused table row. Coordination ablations and event-family decompositions reuse the main coordination pools. Risk-coverage rows reuse the learned-ranking groups. Timing repetitions characterize systems cost and are excluded from the scientific total.
Runtime Environment and Artifacts
Every run records task, seed, candidate count, proposer, active contract, contract outcomes, verifier outputs, selected action, intervention type, outcome labels, and source hashes. Machine-readable manifests link every decision to its candidate-pool digest, task-seed unit, outcome record, and aggregate-table entry.
Claim-Reproduction Order
The minimum reproduction path is:
Verify simulator, proposer, verifier, and configuration revisions.
Materialize the frozen event and natural task-seed matrices.
Generate each ordered candidate pool once and persist its digest.
Run every selector without oracle access and persist its decision.
Label the shared candidate pools and closed-loop executions.
Aggregate paired counts, exact tests, calibration metrics, and denominator checks.
Regenerate the main and appendix tables from the frozen aggregate.
The submitted artifact contains configurations, run manifests, selector records, aggregate tables, statistical scripts, and representative media. Large pretrained proposer weights are referenced by public model revision and checksum rather than duplicated.
Discussion and Limitations
The results support coordination contracts as an interface between prediction and control. Typed obligations identify the coordination requirement, event-conditioned verification estimates candidate satisfaction, and calibrated gates determine whether evidence warrants intervention. Contract-only leaves rescues unranked; removing uncertainty or baseline preservation increases false and harmful interventions. Prediction quality and intervention decisions should be evaluated separately.
Proposal quality and intervention quality form complementary axes. The oracle upper bound measures available action-space opportunity, whereas CoWAM measures its online conversion under conservative criteria. Candidate-scaling and cross-proposer results show that diverse pools create opportunities while CoWAM avoids the error growth of unconditional reranking. Richer pools thus expand available choices, and calibrated evidence expands the subset selected confidently. Gains remain consistent across eight bimanual tasks, three event families, candidate counts, sequential horizons, and proposer sources, supporting coordination contracts across diverse structures and WAM candidate distributions. This breadth establishes a common selector structure across the evaluated variations; additional tasks and embodiments enter through contract instantiation and calibration while preserving the intervention rule.
Evaluation Coverage
The natural evaluation spans held-out seeds across all eight task definitions. The event audit covers synchronization, role, and collision opportunities and connects event-level selection quality to naturally occurring closed-loop coordination outcomes. Cross-proposer evaluation covers X-WAM, LeWorldModel, and mixed candidate pools. Together, these blocks establish selective-intervention gains across tasks, coordination modes, candidate counts, horizons, and proposer sources under one outcome-blind same-pool protocol.
Conclusion
CoWAM uses predicted futures to evaluate coordination contracts before modifying bimanual policy actions. Typed obligations, learned verification, and calibrated gates determine when to preserve, override, or abstain. Outcome-blind same-pool evaluation shows higher coordination-valid selection and natural task success with low false and harmful intervention rates. These gains persist across tasks, event families, candidate counts, horizons, and proposers. Ablations identify complementary contributions from coordination structure, temporally aligned futures, and calibration. Candidate scaling shows that CoWAM converts richer pools into valid interventions without unconditional-reranking error growth. Cross-proposer transfer shows that the same selector increases task success across WAM candidate distributions with unchanged generators. Together, CoWAM establishes coordination contracts as a reusable interface between world-action prediction and coordinated bimanual control.
Frequently Asked Questions
What is a coordination contract in CoWAM? A coordination contract bundles typed admissibility predicates, event-conditioned learned evidence, and calibrated intervention gates that encode synchronization, role, and collision obligations. Every active predicate must pass before an alternative action can override the nominal policy output.
How does CoWAM differ from training a new action generator? CoWAM is purely a selective intervention layer — it never trains a new generator or alters the candidate pool. It consumes the same ordered candidates any selector receives and decides whether to preserve, override, or abstain based on contract evaluation.
What is outcome-blind same-pool evaluation? All selectors see identical candidate pools and must commit their decisions before any simulator outcome labels are available. One shared oracle pass then labels every candidate, preserving an equal information boundary and enabling direct comparison of intervention quality.
What interventions does CoWAM distinguish in its metrics? The evaluation separates rescues (alternatives that repair a failing nominal action), false interventions (changes without task or coordination support), and harmful interventions (changes that lose a required outcome), alongside strict task success and intervention rate.