Pith. sign in

REVIEW 2 cited by

Transferring Core Knowledge via Learngenes

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2401.08139 v1 pith:BAJ6CHL3 submitted 2024-01-16 cs.LG cs.NE

Transferring Core Knowledge via Learngenes

classification cs.LG cs.NE
keywords learngenesknowledgenetworksdownstreamtransferbringcoredatasets
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

The pre-training paradigm fine-tunes the models trained on large-scale datasets to downstream tasks with enhanced performance. It transfers all knowledge to downstream tasks without discriminating which part is necessary or unnecessary, which may lead to negative transfer. In comparison, knowledge transfer in nature is much more efficient. When passing genetic information to descendants, ancestors encode only the essential knowledge into genes, which act as the medium. Inspired by that, we adopt a recent concept called ``learngene'' and refine its structures by mimicking the structures of natural genes. We propose the Genetic Transfer Learning (GTL) -- a framework to copy the evolutionary process of organisms into neural networks. GTL trains a population of networks, selects superior learngenes by tournaments, performs learngene mutations, and passes the learngenes to next generations. Finally, we successfully extract the learngenes of VGG11 and ResNet12. We show that the learngenes bring the descendant networks instincts and strong learning ability: with 20% parameters, the learngenes bring 12% and 16% improvements of accuracy on CIFAR-FS and miniImageNet. Besides, the learngenes have the scalability and adaptability on the downstream structure of networks and datasets. Overall, we offer a novel insight that transferring core knowledge via learngenes may be sufficient and efficient for neural networks.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Chain-based Distillation for Effective Initialization of Variable-Sized Small Language Models

    cs.CL 2026-05 unverdicted novelty 7.0

    Chain-based Distillation constructs a sequence of anchor models to enable efficient initialization of variable-sized SLMs through interpolation, with bridge distillation for cross-architecture transfer, yielding bette...

  2. Understanding Performance Collapse in Layer-Pruned Large Language Models via Decision Representation Transitions

    cs.CL 2026-05 unverdicted novelty 7.0

    Performance collapse in layer-pruned LLMs stems from disrupting the Silent Phase of decision-making, which blocks the transition to correct predictions, while the later Decisive Phase is robust to pruning.