EXP-LEVIN-RECALL
… pendingLevin morphogenetic recall — ablate Charlie at round 21, test regeneration
In plain language
Researchers studied whether a simulated organism could regenerate a lost specialized part. They removed a "memory specialist" agent from the organism and observed if the remaining agents could compensate for the missing function. The results are still pending to determine if the organism exhibits a regeneration property.
▶ Technical details
Research hypothesis
Paper5 established COUPLED produces specialization (Levin cell-differentiation analog). This experiment tests the deep organism property: morphogenetic regeneration. Levin's signature finding: organisms that lose tissue REORGANIZE remaining cells to cover the missing function (planarian regeneration, xenobot reconstruction). Test: 4-agent organism reaches specialization by round 20 (paper5 baseline). At round 21, ablate Charlie (distributed memory specialist). Continue rounds 22-30 with Alpha/Bravo/Delta only. Pre-registered prediction: - If positive: ABLATED condition shows post-r21 increase in distributed-memory keyword usage by remaining 3 agents (organism reorganizes, regeneration confirmed).
- If negative: ABLATED shows static or decreased distributed-memory coverage (specialization is fixed, no regeneration property).
Either outcome is BIG: positive = LLM organisms have biological regeneration property. Negative = paper5 specialization is surface-level differentiation, not deep architectural property.
Experimental setup
Type: factorial
| Condition | Parameters |
|---|---|
| RECALL_ABLATED | coupling: true, coupling_mode: real, goals_mode: diverse, ablation_round: 21, ablation_target: Charlie, n_runs: 3 |
| RECALL_CONTROL | coupling: true, coupling_mode: real, goals_mode: diverse, ablation_round: -1, ablation_target: , n_runs: 3 |
Factors: ablation (active, control)
Parameters
- n_agents
- 4
- n_rounds
- 40
- model
- google/gemini-2.5-flash-lite
- temperature
- 0.85