Model Detail
Darwin-28B-Coder
—Darwin-28B-Coder is a code generation model with 28B parameters released by FINAL-Bench. The model is registered under the text-generation pipeline tag on Hugging Face, distributed under the permissive apache-2.0 license.
Darwin-28B-Coder ships with 28B parameters. Total weight footprint is approximately 26.9 GB, which is the relevant figure when planning local-inference VRAM. The apache-2.0 license is permissive, allowing commercial deployment and derivative work without per-seat fees, though attribution requirements still apply.
Darwin-28B-Coder is best fit for code completion, repository-scale Q&A, and pair-programming integrations. It is a less obvious choice for one-shot generation of security-critical code without review. Treat this as a starting matrix rather than a benchmark verdict — the right deployment usually depends on the specific evaluation suite that mirrors your workload.
Darwin-Claw: Environment-Adaptive Experience Consolidation for Recursively Self-Evolving Agents
arXiv:2610.23574v1 Announce Type: new Abstract: While Large Language Model (LLM) agents have demonstrated remarkable capabilities across complex tasks, they predominantly operate in isolated sessions, lacking the ability to continually learn from historical interactions or adapt to user-specific env
Darwin Family: MRI-Trust-Weighted Evolutionary Merging for Training-Free Scaling of Language-Model Reasoning
arXiv:2605.14386v1 Announce Type: cross 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 re
Direct From Darwin: Deriving Advanced Optimizers From Evolutionary First Principles
arXiv:2605.05284v2 Announce Type: replace-cross Abstract: Evolutionary computation has long promised to deliver both high-performance optimization tools as well as rigorous scientific simulations of Darwinian evolution. However, modern algorithms frequently abandon evolutionary fidelity for physics-
