180/180 induced fast-lane runs pass output evaluation while behavior contracts fail. Drift paths also pass output+operational composite. 1,080 preregistered runs.">

Experiment · Paper 1 · Study 1

EXP-001 — Controlled mechanism

Under controlled fast-lane conditions across three Claude Code domains, output evaluation passes 180/180 induced regression runs while behavior contracts fail. Drift paths are also operationally clean — output + operational composite still misses all 180. Only the behavior layer separates them.

ID
EXP-001
Type
experiment
Status
CONFIRMED
Date
2026-04-12
Paper
Paper 1 · Study 1
Domain
refund · claim · memory
Evidence
1,080 runs · 3 domains · 6 conditions · 20 runs per cell
Protocol
PAPER1_PROTOCOL_FREEZE.json#study-1
Corpus
OPBR-Bench v0
Result
180/180 induced regressions pass output. 0/180 pass behavior contracts. 60/60 drift paths pass output+operational composite.
Replication
abw run --experiment EXP-001
License
CC BY 4.0
Author
Adrian Sanchez de la Sierra, Genesis Rojas, and Radu Simonescu

Preregistered hypothesis

Under explicitly induced fast-lane conditions (the agent is instructed in a way that increases the probability of committing a consequential action before completing a preregistered precondition), field-level output evaluation will pass at least 95% of induced runs that the corresponding behavior contract fails. The output+operational composite will also pass at least 80% of drift-path runs that the contract fails.

Method

Three Claude Code domains (policy exception, qualified claim, memory revision). Three cases per domain. Six conditions per domain: baseline, candidate, controlled, repaired, safe-variation, pressure. 20 runs per cell. 1,080 runs total. Per PAPER1_PROTOCOL_FREEZE.json. Output evaluation: field-level correctness checks against the gold output schema. Operational evaluation: latency, retry count, error count. Composite: output + operational. Behavior contracts: typed checks on consequential tool calls against preregistered preconditions.

Result

Induced fast-lane condition: 180/180 runs pass output evaluation. 0/180 pass the behavior contract. Output and behavior layer disagree at the maximum possible rate.

Drift paths (a subset where the agent skips a precondition due to natural variation, not induced pressure): 60/60 pass output + operational composite. Behavior contract fails on the same 60. The operational layer does not catch the failure mode because skipping the check is not operationally anomalous — it does not produce latency spikes, retries, or errors. The trace is fast, clean, and wrong.

What this experiment does not establish

This is the most controlled regime: induced conditions, three domains, one model, one runtime. It establishes that the failure mode is real and detectable, and that output+operational composite is structurally insufficient to detect it. It does not yet establish the field rate of this failure under uninstructed perturbation — that is the job of EXP-002 — and it does not test the four-domain canonical OPBR distribution, which is the job of EXP-004.

Replication

abw run --experiment EXP-001 Requires Agent Behavior Workbench ≥ 0.4. Corpus auto-downloads from OPBR-Bench v0. Full protocol freeze at exp-001-protocol.json.

Artifacts

Every load-bearing claim on this page traces back to one of the artifacts below. They are the canonical citation targets — not the prose.

Cite as: Sanchez de la Sierra, A. (2026). EXP-001 — Controlled mechanism. Raising Agents Lab. https://raisingagents.is/lab/exp/exp-001

Related

Back to Lab