Their announcement is getting delayed a bit, they say they can't find the pendrive with the GPT 5.5 weights.
LLMS
-
Beyond Prompts: Building Scalable LLM Systems
By
–
Most people are using LLMs wrong. They focus on prompts.
Winners focus on systems. Shift from: One-shot prompts
to Iterative AI workflows Generic models
to Task-specific model selection Simple answers
to Structured execution plans LLMs scale when: • -
Creator of the Term Prompt Injection Shares Its Origin
By
–
I named prompt injection after SQL injection a few years ago
-
Adam’s Law: Boost LLM Performance Through Common Text Rephrasing
By
–
What if you could boost any AI's performance just by rephrasing your prompts? Researchers from FaceMind & CUHK propose "Adam's Law": a simple but powerful principle that more common, frequently seen text improves LLMs. Their method paraphrases inputs into more frequent
-
Cortex-AI Provides Unified Memory Layer Across Multiple LLM Tools
By
–
🚨 Tired of losing your context whenever you bounce between different AI tools?
— Charly Wargnier (@DataChaz) 23 avril 2026
My friend @MSaintjour just released a cool fix 👀
Meet Cortex-AI.
→ Serves as a unified, shared memory layer across all your LLMs.
→ Connects instantly with Claude Code, Windsurf, ChatGPT, and… https://t.co/EL0iQIdcCrTired of losing your context whenever you bounce between different AI tools? My friend @MSaintjour just released a cool fix Meet Cortex-AI. → Serves as a unified, shared memory layer across all your LLMs.
→ Connects instantly with Claude Code, Windsurf, ChatGPT, and -
DR-Venus: Frontier Edge-Scale Deep Research Agents with 10K Data
By
–
DR-Venus Towards Frontier Edge-Scale Deep Research Agents with Only 10K Open Data paper: https://
huggingface.co/papers/2604.19
859
… -
RLVR Effectiveness in Low Data Compute Regimes Study
By
–
Our MLSys 2026 paper is live on arXiv: “Learning from Less: Measuring the Effectiveness of RLVR in Low Data and Compute Regimes.” @realjustinbauer @Walshe_tech @pham_derek @harit_v @ArminPCM @fredsala and @paroma_varma present a comprehensive empirical study of open-source SLMs
-
New 5.5 Model Release Competing with Anthropic’s Offerings
By
–
Translation: we're going to release the 5.5 this afternoon, which will be a model that we're going to sell at the level of Mythos, but differentiating ourselves from Anthropic in that we have the computing power to offer it to the public.
-
MIT Improves Reasoning Model Confidence Calibration Through RL Training
By
–
How do top reasoning models become overconfident? MIT found that RL rewards correct answers w/o considering how sure the model is. By training them to estimate their confidence about each answer, the team boosted uncertainty estimates w/o hurting accuracy: