Wake up, ppl 👀
Xiaomi just open-sourced ~7K of the RL environments it used to train MiMo on HuggingFace, across domains like code, cyber, general, music and web dev
I think this is one of the biggest things to happen in frontier open-source RL environment data.
Indepth analysis and learnings coming soon 👀
I wanted to figure out the best coding agent in open source
So I was recently using opencode and read somewhere to use oh-my-openagent for tokenmaxxing
Then i started using pi and someone suggested to use oh-my-pi for better results
I got amazing results from both of the above, now i am finding out that there is oh-my-codex, oh-my-claudecode also
People in open source are genuinely crazzy
कांग्रेस/SP/BSP आएगी तो जनरल कास्ट को कुछ नहीं होगा।
🔹 RSS बैन होगी…
🔹 मोदी जी जेल जा सकते हैं…
🔹 BJP वालों को सड़क पर दौड़ा-दौड़ाकर मारा जा सकता है…
जनरल कास्ट का क्या है? अभी भी फेक केस में जेल जा रही है, तब भी जाएगी। और हम अपने लिए लड़ लेंगे। 🙏🙏
@BJP4India@BJP4UP
Local Laya moggs Jev at @grok 4.7-built Tetris 🧩
An open-weights System One model called Laya, beat cloud-based Jev at playing Tetris by making decisions 11 times faster, running locally on a 16GB MacBook Air!
Run AI models locally -> https://t.co/RbcCOIgVkj
Jev, now with open weights + vision.
Classified 1,697 SF Tech Week events with Gemma 4 26B-A4B.
Zero labels. No fine-tuning.
Jev-ify any open-source model on SimpleJev.
GPT-6 Astra is by far the best 3D design model today.
& this is THE Complete Tutorial Guide to Design in Codex.
FULL guide you need to know about Codex and GPT-6 to start using it for design work and 3D creative work.
2026: an agent hacked a gym waitlist. Another leaked creds across 4 services. Agent fleets hold real spending power - no gatekeeper.
Built Warden - blocks a bad call live, before it executes. Gemini, ADK, Cloud Run, Vertex AI.
https://t.co/fUXUZievXH
#AllThingsAgenticHackathon
Goons arrive to kidnap Sunny's sister. Sunny flashes his rakhi and shouts his vows to protect her. But just as he is about to beat them to a pulp, the Azaan plays. Now helpless, Sunny sits for Namaz, giving the goons an easy chance to take his sister away.
Pre-social media Bollywood was on another level of delulu.
Insaniyat (1994).
A digital Promise Wall built with Claude Fable 5 in ONE PROMPT using @threejs 👀
I was watching a series the other day where the protagonist runs a beautiful cafe in the mountains.
There was this lovely concept called a Promise Wall where visitors could leave their promises behind.
As an introvert, I randomly thought it would be really cool to bring that idea into so called world of "bites & bytes",... and Claude built it in just a couple of minutes. Crazzzy 🤯
You can add lots of notes to the wall, choose different types, and move around the space with buttery smooth animations. It honestly feels ridiculously satisfying.
So… do you have a promise to make?
Go put it on the wall 🫶
Live: https://t.co/qepXpzNjLV
Code: https://t.co/dzj7BYfriA
Advanced RAG techniques nobody's shipping.
Everyone demos naive RAG.
Almost nobody ships these.
This is the gap between a demo and a product.
1. Contextual Retrieval
Prepend each chunk with an LLM-generated context line before embedding.
Anthropic measured 35% fewer retrieval failures. 67% when combined with reranking.
2. Hybrid Search + RRF
Fuse BM25 (exact terms) with dense embeddings (semantics) via reciprocal rank fusion.
Catches queries that pure vector search misses.
3. Cross-Encoder Reranking
Retrieve 50 fast. Rerank down to 5 with a cross-encoder (BGE, Cohere).
Precision at top-5 jumps dramatically.
4. HyDE
Generate a hypothetical answer first. Embed that not the query.
Fixes vocabulary mismatch between user questions and your docs.
5. Query Decomposition
Split complex questions into sub-queries. Retrieve in parallel. Merge results.
Required for multi-hop questions naive RAG fails on.
6. Small-to-Big (Parent Document)
Embed small chunks for precision. Return the parent chunk for context.
Best of both: tight matching, full context.
7. Late Chunking
Run the whole document through a long-context encoder first. Then pool per chunk.
Every embedding carries global document context.
8. ColBERT / Late Interaction
Token-level matching instead of single-vector similarity.
Wins on technical, domain-specific queries.
9. Semantic Chunking
Split by meaning (embedding breakpoints), not fixed token counts.
Stops chunks from cutting ideas in half.
10. Contextual Compression
Extract only the relevant sentences from retrieved chunks before prompting.
Less noise. Fewer tokens. Higher accuracy.
11. Corrective RAG (CRAG)
Grade retrieved docs. If confidence is low, rewrite the query or hit web search.
Your system self-heals instead of hallucinating.
12. GraphRAG
Extract entities and relations. Build community summaries.
Answers "whole corpus" questions vector search cannot.
13. RAPTOR
Recursive clustering and summarization into a tree. Retrieve at multiple levels.
Multi-hop reasoning across the corpus.
14. Lost-in-the-Middle Ordering
Place strongest docs at the start and end of context. Attention is U-shaped.
Free accuracy gain from ordering alone.
15. Retrieval Evals
Track hit rate, MRR, NDCG on a golden dataset. Block deploys on regression.
You cannot improve retrieval you do not measure.
THE IMPLEMENTATION ORDER
Week 1: Semantic chunking + hybrid search
Week 2: Reranking + parent document retrieval
Week 3: Query decomposition + HyDE for hard queries
Week 4: Corrective loop + evals
Do not ship all 15 at once.
Stack them one at a time. Measure each.
Naive RAG is a demo. Advanced RAG is a product.
One retrieval pipeline plus public benchmarks plus documented tradeoffs
equals more credibility than 10 chatbot wrappers.
Most people ship the tutorial version. Builders ship the production version.
Bookmark & Repost!
Your our history shouldn’t be taught in a boring way 🫠
I’ve always loved history, but I could never imagine what life actually looked like back then.
So I built Empire Atlas, a 3D interactive explorer of 8 historical empires using @threejs and @Kimi_Moonshot K3 🔥
It lets you explore how people lived, what their homes looked like, their maps, daily life, interiors, and more. Properly researched.
But the craziest part is that this was near one shot vibe coded with Kimi K3 🤯
When I previously built a 3D anatomy app with GPT 5.6 sol, I had to iterate on performance and optimization.
With Kimi, the moment I handed over the 3D assets (generated using @tripoai), prompt and design (by GPT Image 2.0), it created an 11 step engineering plan to build the entire thing.
The very first step it did was optimizing the assets.
It took nearly 500MB of 3D assets and brought them down to just 17.8MB using mesh simplification, Draco compression, and 1024px WebP textures.
Absolutely nuts.
It also generated 56 historical images across the 8 empires showing daily life, maps, interiors, and more using its image plugin with batch processing.
Those were converted to WebP too, bringing the total image size to around 10MB.
That’s a huge reason the experience loads so fast on website.
It's engineering workflow or intelligence has really impressed me so far. The only downside is that it took more than 5 hours, though 😅
Anyway, back to history.
In Empire Atlas, you can explore 8 different empires and see how people and our ancestors lived at that time. I really love those textures I was able to create using @tripoai.
You can explore their homes in 3D, and there’s so much more we could do with this.
We could extend these houses into fully explorable interiors and create increasingly realistic reconstructions of what life actually looked like. And maybe create fun education games too.
I genuinely think this can make history education so much more immersive. Much more than showing black and white images in boring textbooks.
Go explore your history now 👇
Live: https://t.co/Dbgxr13L0g
Code: https://t.co/sahbaZeq63
Superior quality in sequence processing can be achieved by entirely removing the complex recurrent and convolutional networks typically used for these tasks.
📄 Attention Is All You Need (Transformer)
https://t.co/BmaidBBuBp
Built by PaperFlakes-zerops
@WeMakeDevs@zeropsio
5M Edge requests. 2K GitHub stars. In just 7 days. 🤯
You all loved this vibe coded project way more than I ever expected. Thank you so much ❤️
To celebrate the milestone, I've added two new features using Claude.
First, interactive quizzes. Instead of simply answering questions in boring style, you now have to locate things on the organ. Where is the aorta? Which one is the left ventricle? A completely new fun way to do quizzes.
Second, multilingual support with RTL support, something many of you have been asking for. For example, when you switch to Arabic, the entire experience automatically shifts from right to left.
I’m planning to add more features, improve the models, and make the experience even more detailed. If you have any feature requests, let me know.
And thank you @sonofalli and @vercel for supporting my lil project ♥️