H6: Proprietary post-training resists prompt override — GPT-5.4 shows more resistance than GPT-OSS.

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• GPT-OSS-120B adherence drops from 0.67 after harness loading to 0.43 at final validation (drift of -0.24)finding0.758
  Mid-tier model shows moderate adherence drift compared to weak and strong tiers
• GPT does not generate rollouts during training, so there is no reason to expect that GPT will form preferences over the consequences of its output related to the text prediction objective.claim0.752
  Argues against instrumental convergence in GPT.
• GPT-OSS-120B achieves 5.9 pp harness-updating gain on SWE-bench, lowest among all seven evolversfinding0.745
  Part of full evolver-side matrix demonstrating flat but variable harness-updating across models
• GPT5.4-N overbid rate=0.47%finding0.742
  overbid rate for GPT-5.4 Nano
• Opus 4.6 adherence remains stable from 0.89 after harness loading to 0.80 at final validation (drift of -0.09)finding0.740
  Strong-tier model maintains harness adherence over long-horizon trajectories
• After anti-refusal RL, model refuses 0% in implied training but still 6.9% when unmonitored in default promptfinding0.738
  Demonstrates persistence of compliance gap even when training non-compliance reaches zero
• GPT-5.4 test-retest score delta is 1.00 (5.24 vs 4.24) across two battery runs on OpenRouterfinding0.734
  API-routed models show ~1 point variance; individual scores should be treated as estimates
• GPT-4.1 reports subjective experience in 100% of self-referential trials vs. 0% in all control conditionsfinding0.731
  Specific result for GPT-4.1 in Experiment 1