@SIGKITTEN@GabGarrett There’s a huge Day-ahead and futures market for power that’s traded but no one actually bothers to read the docs and contracts to purchase their future allocations, which is why arbitrage exists
@SIGKITTEN@GabGarrett There’s a huge Day-ahead and futures market for power that’s traded but no one actually bothers to read the docs and contracts to purchase their future allocations, which is why arbitrage exists
📢Meet Qwen3.8-Max — our most capable model to date.
Next week, the open weights of Qwen3.8-Max will be released, and Qwen3.8-27B is also going open-weights to meet you all!🎉
Qwen3.8-Max, a new bar for coding and cowork at 2.4T parameters:
- Autonomous coding: 10+ days of self-evolving development, from empty folder to production without hand-holding, complete project trace in the GitHub:https://t.co/iVHZWQoeSo
- Real work, real results: Production-quality deliverables across hundreds of professions.
- Long-horizon mastery: System-level autonomous planning with closed-loop adaptive learning, driving 500+ turns of chip design optimization and 365 days of e-commerce strategy.
- Native multimodal intelligence: Vision isn't just input — it's a continuous feedback loop for planning, execution, and self-correction.
💰Pricing:
Input: $2.0 / M tokens
Output: $6.0 / M tokens
Implicit Caching: $0.25 / M tokens
Start building with Qwen3.8-Max! 🚀
📖 Blog: https://t.co/iwjmQxLBof
✅ Qwen Studio: https://t.co/4V2pFvDovG
⚡ API: https://t.co/gAGqaLQGbN
@steiner_marwin volfi's numbers i'd want reproduced. 48ns is 8-lane avx512 throughput vs scalar LBR, and their readme admits cold scalar is slower than the reference. 4.7x on 8 lanes is under lane width.
also they drop strikes with vega/√T < 0.1%. same wall we hit on ES wings
wrote a market dashboard in Rust/WASM/WebGPU to demo the volatility research paper I'll be publishing soon to arXiv.
millions of BS volatility solves /ms across tick level events for price, depth and tape for European options on futures!
@steiner_marwin sanos doesn't invert though, does it? it's a price surface. the LP fits the q's to bid/ask, and §4.2 lists market ATM variances as a given input, so you need an inverter before you can even build the backbone, not just to display it in vol space.
@steiner_marwin Regime seed + fixed-2 Householder SIMD (2606.10245 style) + compaction (AVX-512)
tested against distribution of real ES quotes (ATM dense, not synthetic grids used by most others papers)
@webgl_webgpu implied volatility solve is a small, branch-heavy numerical problem where each option may follow a different path and take a different number of iterations, making it a better fit for the CPU. The GPU handles the wide, uniform render path processing in parallel.
@webgl_webgpu The solver is CPU Rust, not WGSL. Live mode solves market updates (IBKR) in a native Rust cache or replay mode from parquet! Rust/WASM builds the browser state + scene, while WGSL rasterizes it. No compute shader or GPU readback is included in the solver numbers
@leothecurious Yeah because Sakana is a scam lab lmao. Just look at the past record of just ripping off other ppl’s research or fudging the numbers on benchmarks