FULL STORY
Bittensor's Decentralized Training Plagiarism Scandal
Bittensor's Quasar subnet claimed a decentralized 120B model, but was quickly exposed for plagiarism, triggering a trust crisis in the crypto AI space.
2026-07-31 ~ 2026-08-02 · 2 episodes · 5 posts
Episode 1 · Bittensor's Quasar Claims Decentralized Training of 120B Model (2026-07-31, 2 posts)
Bittensor's Quasar subnet claims to have successfully trained a 120-billion-parameter large language model in a decentralized manner, featuring a context window of 5 million tokens. However, the model's capabilities are still pending further verification.
- Bittensor's Quasar Subnet Claims Decentralized Training of 120B Parameter Model — markjeffrey · 2026-07-31
- Bittensor Subnet Claims Decentralized Training of 120B Model with 5M Context, Yet to Be Verified — markjeffrey · 2026-08-01
Episode 2 · Quasar 120B Model Caught Plagiarizing Weights, Token Plummets (2026-08-02, 3 posts)
Bittensor subnet project Quasar faces a severe trust crisis after community members revealed that 96% of its supposedly decentralized 120B model's weights were directly copied from the open-source Ling-mini model, causing its token to crash by 70%.
- Bittensor's Quasar Collapses as 'Trained from Scratch' Model Exposed as 96% Plagiarized — markjeffrey · 2026-08-02
- Quasar Model Accused of Copying Ling-mini Weights, Triggering Bittensor Crisis — markjeffrey · 2026-08-02
- Quasar's 120B Decentralized AI Model Exposed: Weights Match Existing Model, Token Crashes 70% — 0xSammy · 2026-08-02