Read our blog: https://
cerebras.ai/blog/glm-4-7
RESEARCH
-
GLM-4.7 Live on Cerebras: Frontier Coding Intelligence
By
–
GLM-4.7 from @Zai_org is live on Cerebras!
— Cerebras (@cerebras) 8 janvier 2026
– Frontier intelligence for coding, tool-driven agents, and multi-turn reasoning
– Record coding speed: ~1,000 tokens per second (up to 1,700 TPS for other uses)
– Strong price-performance: ~10x higher than Sonnet 4.5 pic.twitter.com/HGTcVvhZVQGLM-4.7 from @Zai_org is live on Cerebras! – Frontier intelligence for coding, tool-driven agents, and multi-turn reasoning
– Record coding speed: ~1,000 tokens per second (up to 1,700 TPS for other uses)
– Strong price-performance: ~10x higher than Sonnet 4.5 -

Interview on Optuna History: From 2018 Challenges to Today
By
–
Honored to be interviewed about the history of Optuna. It covers everything from our early challenges in 2018 to the mindset I carry into my work today. Full interview (in Japanese): findy-code.io/engineer-lab/i… Takuya Akiba (@iwiwi) Optunaの開発秘話についてインタビューして頂きました! 2018年当時のOptuna誕生の経緯と裏側、この経験の振り返りや学び、今の仕事にも繋がる自分の根底にある考え方、などを上手く引き出してまとめてくださってます。ぜひ! findy-code.io/engineer-lab/i… — https://nitter.net/iwiwi/status/2009079954907213846#m
→ View original post on X — @_yutaroyamada, 2026-01-08 17:00 UTC
-
OpenTinker: Train AI Agents on Laptop Easily
By
–
What if you could train AI agents on a laptop as easily as on a GPU cluster?
— 机器之心 JIQIZHIXIN (@jiqizhixin) 8 janvier 2026
Researchers from UIUC's U Lab, led by Prof. Jiaxuan You, just open-sourced OpenTinker.
It's a new "Reinforcement-Learning-as-a-Service" (RLaaS) system that decouples the complex training pipeline into… pic.twitter.com/YSH2txDsEhWhat if you could train AI agents on a laptop as easily as on a GPU cluster? Researchers from UIUC's U Lab, led by Prof. Jiaxuan You, just open-sourced OpenTinker. It's a new "Reinforcement-Learning-as-a-Service" (RLaaS) system that decouples the complex training pipeline into
-
Digital Red Queen: Adversarial Program Evolution in Core War with LLMs
By
–
Digital Red Queen: Adversarial Program Evolution in Core War with LLMs
— hardmaru (@hardmaru) 8 janvier 2026
Blog Post: https://t.co/DwllY5YNOD
Full Paper (Website): https://t.co/PKfXJTWfRN
Full Paper (arxiv): https://t.co/V0elSPvSaQ
More information:https://t.co/GD1iLlvnUCDigital Red Queen: Adversarial Program Evolution in Core War with LLMs Blog Post: https://
sakana.ai/drq/
Full Paper (Website): https://
pub.sakana.ai/drq/
Full Paper (arxiv): https://
arxiv.org/abs/2601.03335 More information: -

Dynamic Context Discovery in Large Language Models
By
–
Dynamic context discovery · Cursor https://
buff.ly/xpdnlPc
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
Core War: Adversarial Program Evolution and Digital Red Queen
By
–
Survival of the fittest code.
— hardmaru (@hardmaru) 8 janvier 2026
Core War (1984) is a game where programs must crash their opponents to survive. Warriors written in an assembly language called Redcode fight for control of a virtual machine.
Our new paper: Digital Red Queen: Adversarial Program Evolution in Core… pic.twitter.com/f5XGxfCJCFSurvival of the fittest code. Core War (1984) is a game where programs must crash their opponents to survive. Warriors written in an assembly language called Redcode fight for control of a virtual machine. Our new paper: Digital Red Queen: Adversarial Program Evolution in Core
-
Imaginary RoPE Implementation Explored for Language Models
By
–
Check out the imaginary RoPE implementation here!
-

RoPE++ Improves Long-Range Dependencies in Language Models
By
–
The "imaginary" part of RoPE contains valuable information for long-range dependencies RoPE++ reintroduces it in heads with slower decay curves that better capture distant tokens We provide implementations with one-command scripts for both the original codebase and Nanochat!