question
active
question:how-can-persona-reweighting-be-mechanistically-distinguished-from-persona-model-collapseHow can persona reweighting be mechanistically distinguished from persona-model collapse?
Central open problem identified by the authors: the same mechanistic signatures may be consistent with both accounts
Source paper
extracted_from(2026) · Davi Bastos Costa · Renato Vicente
Neighborhood — ranked by edge-count
Papers (1)
paper
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Authors argue collapse is a separate process from the reweighting account, not merely a relabeling
- Authors argue the mechanistic evidence typically cited for reweighting is equally consistent with their collapse account
- Proposed mechanism for collapse distinct from reweighting: representation bleeding rather than archetype selection
- Core definitional sentence for the paper's central concept
- Extended experimentation proposed to clarify the extent of the findings
- The process by which fine-tuning shifts model behavior toward dark character archetypes already present in pre-trained repertoire
- Is persona-model collapse gradual or sudden during fine-tuning, and does it track standard training-loss signals?question0.815Future direction: monitoring S and R over the course of fine-tuning could reveal collapse dynamics
- The central concept introduced by this paper: deterioration of a model's internal capacity to simulate, differentiate, and maintain consistent characters