ByteShape

@ByteShape

Accelerating AI by Learning Optimal Datatypes

Canada

byteshape.com

Joined October 2024

7 Following

180 Followers

Replies Reposts Media

ByteShape
@ByteShape
Oct 25
ByteShape is at #EMNLP2026 in Budapest! Come see our Industry Track poster, "Displacement Is Not Direction", in Poster Session D! 📍 Hall H, HUNGEXPO 📅 Mon, October 26, 11:00–12:30 Can't make it? Find us anytime October 25–29! 📄 arxiv.org/abs/2606.19558 #NLProc #ByteShape
arxiv.org
Displacement Is Not Direction: Evaluating Fidelity Metrics for...
Fidelity metrics, such as per-token KL divergence (KLD) against a high-precision reference, are often used in practice as low-cost proxies for benchmark quality. We test this practice on a...
ByteShape
@ByteShape
Oct 19
ByteShape is a Startup Sponsor at #PyTorchCon 2026! Come visit us at Booth #52 for fun interactive demos, conversations about efficient AI, and a special surprise! 📸 Take a photo and bring home a printed keepsake! 📍 San Jose Convention Center 📅 October 20–21 See you there!
ByteShape
@ByteShape
Sep 18
1/3 We updated our Qwen3.8 benchmarks to include Prism-LM’s new Bonsai 2 QAT models. These are our results, not the benchmark numbers reported by Prism. We ran Bonsai 2 using Prism’s runtime, while keeping the rest of the setup consistent with our other Qwen3.8 evaluations.
8
ByteShape
@ByteShape
Sep 15
We released ShapeLearn-Lite days after Qwen 3.8 27B, and it already delivered some of the best quality/speed trade-offs we tested. Now full ShapeLearn is even better: 99.63% of BF16’s aggregate benchmark score at just 3.84 bpw. Models & results: byteshape.com/blogs/Qwen3.8-…
2
ByteShape
@ByteShape
Aug 12
We recently released two optimized image-generation models: Qwen-Image-2512 and Krea-2-Turbo. We pushed both into the ~3 bpw range while preserving their ability to generate high-quality images. Each model comes in two flavors: (1/2)
1