Episode
Darwin Family: MRI-Trust-Weighted Evolutionary Merging for Training-Free Scaling of Language-Model Reasoning
- Podcast
- Daily Paper Cast
- Published
- May 16, 2026
- Duration seconds
- 1402
- Processing state
not_requested- Canonical source
- https://share.transistor.fm/s/0f4c1b96
Actions
POST https://stenobird.com/v1/public/podcasts/daily-paper-cast-7079649/episodes/darwin-family-mri-trust-weighted-evolutionary-merging-for-training-free-scaling-of-language-model-reasoning/transcription-requests
Idempotently request low-priority transcript generation for this episode.GET https://stenobird.com/podcast/daily-paper-cast-7079649/darwin-family-mri-trust-weighted-evolutionary-merging-for-training-free-scaling-of-language-model-reasoning.md
Read the agent-friendly Markdown representation of this episode resource.
Summary
🤗 Upvotes: 44 | cs.NE, cs.AI Authors: Taebong Kim, Youngsik Hong, Minsik Kim, Sunyoung Choi, Jaewon Jang, Junghoon Shin, Minseo Kim Title: Darwin Family: MRI-Trust-Weighted Evolutionary Merging for Training-Free Scaling of Language-Model Reasoning Arxiv: http://arxiv.org/abs/2605.14386v1 Abstract: We present Darwin Family, a framework for training-free evolutionary merging of large language models via gradient-free weight-space recombination. We ask whether frontier-level reasoning performance can be improved without additional training, by reorganizing latent capabilities already encoded in existing checkpoints. Darwin introduces three key ideas: (i) a 14-dimensional adaptive merge genome enabling fine-grained component- and block-level recombination; (ii) MRI-Trust Fusion, which adaptively balances diagnostic layer-importance signals with evolutionary search through a learnable trust parameter; and (iii) an Architecture Mapper that enables cross-architecture breeding between heterogeneous model families. Empirically, the flagship Darwin-27B-Opus achieves 86.9% on GPQA Diamond, ranking #6 among 1,252 evaluated models, and outperforming its fully trained foundation model without any gradient-based training. Across scales from 4B to 35B parameters, Darwin models consistently improve over their parents, support recursive multi-generation evolution, and enable a training-free evolutionary merge that combines Transformer- and Mamba-based components. Together, the Darwin Family demonstrates that diagnostic-guided evolutionary merging is a practical and reproducible alternative to costly post-training pipelines for reasoning-centric language models.