Rohit Kumar Tiwari

@analyticalrohit.bsky.social

Simplifying AI for Everyone | Sr. Data Scientist | Creator @ AwesomeNeuron | Top 1% @ Topmate | Writes to 20K+ @ LinkedIn | GenAI, LLMs, Agents Join 10K+ AI Folks → awesomeneuron.substack.com

GLM-5.2 is the strongest open model to date. Can you really run GLM 5.2 locally? GLM 5.2 has 744B params (40B active) and a 1M context window. The 2 bit model retains ~82% accuracy after being shrunk from 1.51TB to 238GB an ~84% reduction in size Run on a 256GB Mac or RAM/VRAM setups

Bild

GLM-5.2 is the strongest open model to date. Can you really run GLM 5.2 locally? GLM 5.2 has 744B params (40B active) and a 1M context window. The 2 bit model retains ~82% accuracy after being shrunk from 1.51TB to 238GB an ~84% reduction in size Run on a 256GB Mac or RAM/VRAM setups

Bild

The best site on the internet for Harness Engineering, completely free. The better the harness, the better the agent. AI coding agents are powerful but without the right controls. > they can be inconsistent > make mistakes > struggle with larger tasks That is where Harness Engineering comes in.

Bild

The best site on the internet for Harness Engineering, completely free. The better the harness, the better the agent. AI coding agents are powerful but without the right controls. > they can be inconsistent > make mistakes > struggle with larger tasks That is where Harness Engineering comes in.

Bild

Mastering these NLP papers places you in the top 1% in AI. The MOST Influential: GPT, BERT, Attention, DeepSeek. - Word2Vec - BERT - GPT-2 - GPT-3 - Attention is All You Need - Mixture of Experts (MOE) - DeepSeekMath (with GRPO)

Bild

Mastering these NLP papers places you in the top 1% in AI. The MOST Influential: GPT, BERT, Attention, DeepSeek. - Word2Vec - BERT - GPT-2 - GPT-3 - Attention is All You Need - Mixture of Experts (MOE) - DeepSeekMath (with GRPO)

Bild