Community open-sources theoretical reconstruction of Claude Mythos
Shruti_0810 · x · 2026-08-27
A community-driven project called OpenMythos has been open-sourced, providing a theoretical reconstruction of Anthropic's Claude Mythos architecture. Built as a Recurrent-Depth Transformer, it features:
- Recurrent reasoning blocks
- Switchable MLA / GQA attention
- Sparse MoE feed-forward layers
- Configurable reasoning depth
- PyTorch implementation
The author clarifies this is not an official Anthropic release but a community implementation based on public research and speculation. It serves as a valuable resource for exploring how such advanced architectures might function under the hood.
Related event: OpenMythos: Open-Source Theoretical Recreation of Claude Architecture(3 posts)→
More from Models
- New inference engine 'e' hits 10,500 tok/s on a single DGX Station — brandon_galang · 2026-10-16
- Reasonable post-trains Nemotron 3.5 on 3,000 Verus proofs, lifting pass@1 ~30 points to 73.3% — fhuszar · 2026-10-16
- NVIDIA's Nemotron 3.5 Super VL hits CoreWeave: 12B-active multimodal MoE at $0.25/M input tokens — wandb · 2026-10-16
- Opus 5 might be a very deep model or use looping, commentator speculates — scaling01 · 2026-10-16
- GPT-6.1 Sol audit finds 6% of arXiv papers had false main theorems, down to 1.5% recently — nitarshan · 2026-10-16
- Nvidia-backed post-training pushes Nemotron 3.5 VL past frontier on dermatology diagnosis — LoganGrasby · 2026-10-16