@philippmusfeld.bsky.social , @joschadutli.bsky.social, @koberauer.bsky.social and I wrote up the result of three years of in lab discussions on setting priors in GLMs.
Hope some of you find the proposed workflow and our recommendations helpful!
#bayes #glm #brms #rstats
pony-alpha-2 ranked the reviews in this order:
1. pony-alpha-2 — best fixes, good coverage, practical
3. claude — widest coverage but noisy
3. glm5 — missed the biggest bug, some incorrect fix suggestions
Very interesting!
#glm #Zai #LLM #Opus #Claude
株式会社グローバル・リンク・マネジメント、再生事業戦略を発表し新たな挑戦を開始 #東京都 #文京区 #GLM #再生事業 #パシフィックスクエア千石
株式会社グローバル・リンク・マネジメントが、東京都文京区で再生事業戦略発表会を開催。中古物件の付加価値向上に挑む新たな取り組みを紹介。
#statstab #501 Prediction intervals for GLMs
Thoughts: Sometimes the prediction of the next dat point can be [0,1]. Not very useful.
#prediction #uncertainty #predictionintervals #glm #binomial #probability
fromthebottomoftheheap.net/2017/05/01/g...
Больше моделей, больше возможностей: зачем мы вводим подписки в Kodacode У нас появились долгожданные подписки!...
#kodacode #модели #llm #искусственный #интеллект #codex #gpt-5.3-codex #glm
Origin | Interest | Match
Building the future, one block at a time. 🏗️📐#CodingCommunity #Tech #GLM #Developers #AiCoding
GLM 5.0 Coding Plan: dub.sh/glm5
Whether you're refining your logic or starting fresh, this plan is built to move as fast as you do.
Grab 10% off with code MCERSXR9BE.
Valid through March 2026.
Explore new #Python versions, upcoming `frozendict`, #Claude 4.6, #GPT 5.3, #GLM 5, #Agent Skills and more fun stuff in the February edition of "This Month for Pythonistas" blueberrypy.netlify.app/post/this-mo...
Your AI girlfriend is nothing but a GPU, either in your gaming rigs or in some datacenter out there in the world at the end. Deal with it 😎
February's about to end tomorrow so here's to all the fellow lonely male people out there in the world lololol :p
#ai #llm #aiSlop #stablediffusion #chatgpt #grok #claude #mistral #qwen #deepseek #glm #kimi #flux
The #LLM job of converting the AI framework was not trivial and would have required me studying lots of machine learning stuff to understand what I would inevitably got wrong on a first attempt at running the OCR model. Pair programming with #kimi and #glm was surprisingly good.
Qwen3.5-35B-A3B vs GLM-4.7-Flash: Two Chinese MoE Models, Very Different Strengths
awesomeagents.ai/tools/qwen-3-5-35b-a3b-v...
#Qwen #Qwen35 #Glm
"Is ANOVA relevant?" My latest column article in WII e-newsletter looks at the historical dichotomy of #ANOVA vs #Regression in #biostatistics classes, while both are mathematically similar.
Read- tinyurl.com/yxnxmwm4
#Statitcs #QuantitativeEcology #GLM #ResearchMethods #AcademicWriting #Ecology
Yesterday, #OpenAI released GPT-5.3-Codex-Spark as a research project for Pro users. With the big announcements coming from China with the releases of #GLM 5, #MiniMax 2.5, plus the hype #Seendance and #Kling are generating, it might seem less relevant.
#ChatGPT #GenAI #LLM #dev #developer #SWE #AI
In my tests I didn't find GLM 5 thinking at Opus 4.5 level, slightly (subjective) better than Kimi K2.5 thinking I'd say.
GLM 5 is included in the $8 sub at nano!
nano-gpt.com/subscription...
#glm #opus #nanogpt
Zhipu #AI veröffentlicht mit #GLM-5 ein Open-Source-Modell mit 744 Mrd. Parametern. In Benchmarks soll es nahe an #Claude Opus 4.5 und GPT-5.2 heranreichen. Das MoE-Modell steht unter MIT-Lizenz und unterstützt auch chinesische Chips. Parallel erscheint das RL-Framework slime.
An interesting weeks after we've got Opus 4.6 and GPT 5.3-Codex, now, we have GLM 5
GLM-5: From Vibe Coding to Agentic Engineering share.google/qeUY6pkWMv1X...
#AI #LLM #GLM
#statstab #479 Different ways of calculating OLS regression coefficients (in R)
Thoughts: The are many ways to skin a variable...
#rstats #regression #modelling #tutorial #r #glm #ols #coding
thomvolker.github.io/blog/2506_re...
El tema hoy con los #SOTA esta candente
Por un lado #Anthropic lanzando #opus 4.6 (meh)
Por otro #openAI lanzando #GPT 5.3 #codex (mas meh)
Mientras, yo dandole cariño a kimi y probando en local #GLM
Global Link Management Announces Fiscal Year 2025 Financial Results Meeting for Investors #Japan #Tokyo #Global_Link_Management #GLM
Been working on something for terminal lovers 👨💻
Codeep – AI #coding agent that lives in your terminal
✓ #GLM coding plan
✓ #MiniMax coding plan
✓ More LLMs coming soon 🔜
#AI #DevTools #Terminal #BuildInPublic #OpenSource
Deep into Code:
vladoivankovic.com/Project?id=c...
ClaudeCode에서 GLM 모델의 실시간 사용량을 한눈에 확인할 수 있는 statusline 플러그인. 📊
✅ 토큰 사용량 실시간 모니터링
✅ 5초마다 자동 갱신
이제 터미널 하단에서 GLM 모델 사용량 걱정 없이 코딩에 집중하세요! 💻✨
🔗 github.com/jeongsk/g...
#ClaudeCode #GLM #Zai
Events though I now put it clearly in my AGENTS.md file, my #ai #agent fails using the constants used in the project and continues to hardcode values. Every review adds a commit titled 'no hardcoded values' [1] ! What more can I do?.... #llms #opencode #glm #rust
1: github.com/asfaload/asf...
Zhipu drops GLM-4.7-Flash: 30B-MoE, 3B-hot, MLA from DeepSeek, 59.2 SWE-bench, free API, M5-local at 43 tok/s—will it become your go-to code sidekick?
#Zhipu #GLM #DeepSeek
open.substack.com/pub/aidisrup...
#AI doesn't help you think 🤔 better.
#AI amplifies 📣 whatever you bring to it.
We are living through the most significant ♾️ transformation in human cognition since the invention of ✍ writing.
open.substack.com/pub/seldonda...
#ChatGPT #Claude #Gemini #Grok #Mistral #Deepseek #Kimi #GLM #Minimax
#llm #GLM #CodingJourney #agent (By きしだのHatena )2026-01-06 GLM-4.7で自宅コーディングエージェントが現実的に。日本語力も高く幅広く使える。 - きしだのHatena share.google/RIXDa2MWkhCY...
→ 100万円で確実にLLMサポートしてもらえるコーディング環境が作れそうですが、個人で考えるとそれなりの出費ですが、稼働年数から人雇うより安いかも？ 環境も作れるので。
Тестирую ChatGPT, Claude, DeepSeek, Grok и ещё 5 нейросетей на реальном запросе: кто поможет активировать Windows? Друзья, все...
#chatgpt #claude #deepseek #grok #gemini #gigachat #qwen #glm #нейросети #ии
Origin | Interest | Match
Come and register for our cloud platform !🙌 You can access various open models via API calls.
Click my link 👉https://cloud.canopywave.io/referral/88DBFXX8G6 → sign up → $40 free credits – 30 seconds activation 🥳
#DeepSeek V3.2, #KimiK2Thinking, #GLM 4.7 all live
