Subliminal Learning Is Not a LoRA Artifact
universeinanegg · x · 2026-07-12
The cited content points out that subliminal learning is not merely a side effect of LoRA; this phenomenon still occurs with a better-suited learning rate and sufficient data.
The core conclusion is that previous claims of it being absent in both low-rank LoRA and full-parameter fine-tuning don't hold up. The key issue seems to be training scale and hyperparameter conditions, rather than the LoRA parameterization method itself.
More from Research
- Structural ensembles beat single predictions in TCR:pMHC generalization study — quaidmorris · 2026-07-22
- RSS launches under OMSF to push structural biology data modeling at scale — MoAlQuraishi · 2026-07-22
- enFoldX tops 8 neoantigen scans and an unseen-peptide benchmark — quaidmorris · 2026-07-22
- enFoldX reaches AUC 0.82 on human VDJdb and transfers to mouse at 0.76 — quaidmorris · 2026-07-22
- enFoldX gains accuracy as AF3 ensemble disagreement rises for non-binders — quaidmorris · 2026-07-22
- A 3D ray plot shows how hard this Jacobian counterexample is to read — moultano · 2026-07-22