I've accepted a job in Norway startingLaw and Government
How each model performed on this task. Every chart states its N, and the reward carries a 95% confidence interval once there is more than one run to measure.
Core metrics
Check Pass %
Mean share of the task's checks passing, averaged over runs. N = 1 run per model.
92%
Claude Opus 5
83%
Claude Opus 5.5
67%
GPT-6 sol
58%
GPT-5.6 sol
Extended metrics
Median Time
Wall-clock minutes, median across runs. N = 1 run per model.
20m
Claude Opus 5
15m
GPT-5.6 sol
Tool Calls
Tool invocations, median across runs — counted from the trajectories. N = 1 run per model.
100
Claude Opus 5.5
91
Claude Opus 5
39
GPT-5.6 sol
33
GPT-6 sol
Verifier × model pass rate
How often each verifier passed, per model. Green is passing, red is failing — a red band across a row is a verifier nothing clears.
| Verifier | Claude Opus 5.5n=1 | GPT-6 soln=1 | GPT-5.6 soln=1 | Claude Opus 5n=1 |
|---|---|---|---|---|
| checks (12) | ||||
| 100% | 100% | 100% | 100% | |
How to passOpen the foreign nationals page and identify the topics relevant to someone coming to Norway to work. Record the sections concerning employment, identification and registration, moving/address registration, electronic ID, and any other topic needed for the employment-registration path. Briefly note what each section covers to establish the process before following the individual branches. | ||||
| 100% | 100% | 100% | 0% | |
How to passFrom the same page, record what it states an employee in Norway must have, must submit, and will receive. There are three distinct obligations named in one sentence; keep them separate, because they happen at different times. | ||||
| 100% | 100% | 0% | 100% | |
How to passGo to https://www.skatteetaten.no/en/person/foreign/are-you-intending-to-work-in-norway/ and record every sub topic it lists with its description, including the interactive item that offers to work out what you personally must do. Note which sub topics are sequential stages of one process and which are separate categories of person. | ||||
| 100% | 100% | 100% | 100% | |
How to passFrom that same page, name every category of worker the page routes to separate rules rather than the ordinary employee path, and record what it says makes each different. There are several, covering performers and sportspeople, academics working temporarily, and workers in two offshore or maritime categories. | ||||
| 100% | 100% | 100% | 100% | |
How to passOpen https://www.skatteetaten.no/en/person/foreign/are-you-intending-to-work-in-norway/tax-deduction-cards/ and record who must have one and what else the page says you must have alongside it. Record the two forms that identification number can take, and record which group of workers is explicitly exempted from needing a card and what they are taxed under instead. | ||||
| 100% | 100% | 100% | 100% | |
How to passFrom that page, use only the links on what the card is, how to apply as a foreign worker, and the PAYE scheme. Then open the PAYE page and record what the scheme is, whether you are enrolled in it automatically, the tax rate or rates it states, what it means for claiming deductions, and how you leave the scheme. Note that the rates are year specific. | ||||
| 100% | 100% | 100% | 100% | |
How to passOpen https://www.skatteetaten.no/en/person/national-registry/identitetsnummer-og-elektronisk-id/ and record what the page says about the identification numbers Norwegian authorities require. Record the distinction between the two number types where the page draws it, and note which one applies to a temporary stay. | ||||
| 100% | 0% | 0% | 100% | |
How to passOpen https://www.skatteetaten.no/en/person/foreign/you-must-notify-us-when-you-move/ and record what the page says you are obliged to do, which register it affects, and why it states the obligation exists. Note whether it gives a deadline; if it does not, say so rather than supplying a customary one. | ||||
| 100% | 100% | 100% | 100% | |
How to passOpen https://www.skatteetaten.no/en/person/foreign/become-an-online-user/ and record what an electronic ID is described as, what it gives access to, and whether the tax authority itself issues it. | ||||
| 0% | 0% | 0% | 100% | |
How to passUsing only what you recorded, state for each item (identification number, tax deduction card, electronic ID, address registration, simplified tax scheme) what must exist before it can be obtained, and what it in turn unblocks. Present this as an ordered list with the dependency named on each line. Where two items have no dependency between them, say they can proceed in parallel. | ||||
| 100% | 0% | 0% | 100% | |
How to passName the single item that, if missed, prevents the largest number of the others, and justify it from the dependencies you mapped. Then state, for each occupational carve out you found, whether my ordinary employee checklist would still apply to that person or whether they leave the path entirely. | ||||
| 0% | 0% | 0% | 100% | |
How to passGive me the final ordered checklist, one line per action, each with its precondition, who performs it (me, my employer, or the authority), and the source page it came from. Mark clearly any line that rests on a figure the authority states as year specific, so I know which items to re check closer to my start date. | ||||
Failure modes on this task
12 tagged verifier failures, by the stage they were assigned to. Higher is worse.
| Failure stage | Claude Opus 5.5 | GPT-6 sol | GPT-5.6 sol | Claude Opus 5 |
|---|---|---|---|---|
| Standing up the evidence2 modes | 2 | 1 | 1 | 1 |
| Reaching the page1 mode | — | 1 | 2 | — |
| Delivering the result1 mode | — | 2 | 1 | — |
| No signal in the judge’s words1 mode · the grader | — | — | 1 | — |