FA-100601 / Mixture of experts dispatch / Open access

Arrival order wins capacity over stronger routes · case 01

Arrival order wins capacity over stronger routes.

Verified by executionVariant 1 · 6 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

Capacity admission ignores the contract’s score priority.

THE FAILURE

Capacity admission ignores the contract’s score priority.

Unsuccessful approach: Global score sorting admits the right rows but disrupts token order.

Case contract

Choose up to capacity rows per expert by score descending then token ID; return selected rows in original order.

Why this case matters

A deterministic toy router makes this dispatch decision independently inspectable.

1 / The failure

Exit 1

"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(routes,capacity):
    return routes[:capacity]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
e='e'+str(N); rows=[[0,e,1],[1,e,9],[2,e,5]]
check('best two in arrival order',solve(rows,2),[rows[1],rows[2]])
check('best one',solve(rows,1),[rows[1]])
check('all fit',solve(rows,5),rows)
check('zero',solve(rows,0),[])
check('empty',solve([],N),[])
check('per expert quota',solve([[0,e,1],[1,'x',2],[2,e,3]],1),[[1,'x',2],[2,e,3]])
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)

Boundary fixture	Actual	Expected	Outcome
best two in arrival order	`[[0, 'e1', 1], [1, 'e1', 9]]`	`[[1, 'e1', 9], [2, 'e1', 5]]`	Failed
best one	`[[0, 'e1', 1]]`	`[[1, 'e1', 9]]`	Failed
all fit	`[[0, 'e1', 1], [1, 'e1', 9], [2, 'e1', 5]]`	`[[0, 'e1', 1], [1, 'e1', 9], [2, 'e1', 5]]`	Passed
zero	`[]`	`[]`	Passed
empty	`[]`	`[]`	Passed
per expert quota	`[[0, 'e1', 1]]`	`[[1, 'x', 2], [2, 'e1', 3]]`	Failed

SHA-256 / 94b3c5fe6085da96117dbd77dd1b9c160a2c81caed418703056530a53abb8e84

2 / The unsuccessful fix

Exit 1

"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(routes,capacity):
    return sorted(routes,key=lambda r:(-r[2],r[0]))[:capacity]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
e='e'+str(N); rows=[[0,e,1],[1,e,9],[2,e,5]]
check('best two in arrival order',solve(rows,2),[rows[1],rows[2]])
check('best one',solve(rows,1),[rows[1]])
check('all fit',solve(rows,5),rows)
check('zero',solve(rows,0),[])
check('empty',solve([],N),[])
check('per expert quota',solve([[0,e,1],[1,'x',2],[2,e,3]],1),[[1,'x',2],[2,e,3]])
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)

Boundary fixture	Actual	Expected	Outcome
best two in arrival order	`[[1, 'e1', 9], [2, 'e1', 5]]`	`[[1, 'e1', 9], [2, 'e1', 5]]`	Passed
best one	`[[1, 'e1', 9]]`	`[[1, 'e1', 9]]`	Passed
all fit	`[[1, 'e1', 9], [2, 'e1', 5], [0, 'e1', 1]]`	`[[0, 'e1', 1], [1, 'e1', 9], [2, 'e1', 5]]`	Failed
zero	`[]`	`[]`	Passed
empty	`[]`	`[]`	Passed
per expert quota	`[[2, 'e1', 3]]`	`[[1, 'x', 2], [2, 'e1', 3]]`	Failed

SHA-256 / 590ebd86c3ff6980b92194f6122796861aa53d342fe689d4c1da8b451415f2cb

HELD IN THE MEMBER ARCHIVE

The verified repair and its recorded checks are member-only.

This mechanism has 6 recorded checks per implementation. The open-access tier publishes the failure and the unsuccessful fix; the repaired source that passes every check, and the observations that prove it, are available to members.

Every case sharing this mechanism uses the same contract and the same repair, so this one record is held back for all of them.

Member access is invitation-based. Sign in with your invited account to inspect the repair.

Verification & scope

Inputs are finite toy scores and routes; the contract does not model a production model, accelerator, or learned routing quality. This reproducer isolates one failure mechanism. Results cover the supplied fixtures. Variants within a family share a test contract and should remain grouped when constructing evaluation splits. Related mechanisms with a shared evaluation_group must also remain together; these controlled models are not independent production incidents.

Observations recorded using Python 3.12.14 at 2026-09-30T16:48:37.615583+00:00.

Case digest / 0fa74b37942024df63ac904b058810b5db3cf41a632d1bd28c858b29cc3bd9eb