QuantFather
28K posts
@QuantFather63
SN63 🤌 $TAO 🤖
This is Quasar Attention, the mechanism behind the upcoming Quasar models, designed to support context lengths of up to 5 million tokens. Attention has long been a bottleneck for processing extended context. Standard attention mechanisms struggle to scale beyond ~200k tokens in training, creating a ceiling on how much information models can reliably use. One approach to solving this has been linear attention methods, such as gated delta attention (used in Qwen 3.5) or Kimi delta attention. These improve efficiency and allow longer sequences, but introduce trade-offs: instability at extreme lengths, quality degradation, and in practice, they are not strictly linear. Quasar Attention takes a different approach. It uses a continuous-time formulation, implemented as a fully matrix-based system rather than relying on vector-state approximations. In practice, this improves stability, reduces cost, and maintains performance as sequence length increases. In internal stress tests at 50 million tokens, KDA-based approaches begin to lose stability, while Quasar Attention remains stable. This allows performance to hold as sequence length increases, rather than degrading beyond a fixed threshold. On BABILong, a Quasar-based model pretrained on 20B tokens and fine-tuned on 16k sequences was evaluated on contexts ranging from 1 million to 10 million tokens, maintaining consistent performance across that range. By contrast, models using gated delta attention show significant degradation at longer lengths, in some cases dropping to ~10% performance at 10 million tokens. (Note: results are indicative; setups are not directly comparable) On RULER benchmarks, a Quasar-10B model (built on Qwen 3.5 with frozen base weights and Quasar Attention added), pretrained on 200B tokens, achieved 87% at 1 million tokens, outperforming significantly larger baselines, including Qwen3 80B, under the same evaluation conditions. Taken together, this points to a shift in where long-context performance is won or lost: not in model size alone, but in the attention mechanism itself. Quasar Attention represents a step change in long-context modelling, setting a new standard for stability and performance at scale. We thank @TargonCompute for the compute and for being our compute provider and long-term partner in training the upcoming Quasar models Here is the link to our paper 👇
Bittensor will be run by agents. They will feed the mining, resist the exploits, manage the fleets, build the subnets and consume the commodities
Bittensor's biggest subnet receives $52M a year in TAO emissions. It generates $2.4M in actual revenue. Without the subsidy, it would cost more than AWS.
JUST IN: OpenAI reportedly offering private-equity firms a guaranteed minimum return of 17.5% in push to raise fresh capital.
Which Bittensor $TAO subnet under $5M market cap are you most bullish on and why?
Closing in on 3rd place on the world-renowned AEC benchmark. Bitcast subscribers up 35% in 20 days. AEC = Attention Economy of Crypto Completely made up benchmark. Real growth.
Big level coming up on $tao weekly level with htf 0.5 fib.
Yuma's Large Cap Subnet Fund is up 76% YTD on a USD basis; 46% higher than $TAO/USD growth over the same period. The growth is driven by targeted exposure to Bittensor's largest subnets like @tplr_ai (SN3), @TargonCompute (SN4), and @webuildscore (SN44), among others.
