Top all-time · #5470
blog.normalcomputing.ai
Company blog · blog.normalcomputing.aiaihardware
blog.normalcomputing.ai is the company blog at blog.normalcomputing.ai, writing about ai and hardware. Since 2023, 14 different Hacker News readers have linked it 14 times, most often “Supersizing Transformers: Going Beyond Rag with Extended Minds for LLMs”. No feed was found, so read it on the site itself.
Too few sampled comments (6) to say how often HN recommends it.?
Why HN links it
HN readers link this blog for posts on getting LLMs to produce constrained output, such as regex-guided valid JSON, and on grounding LLMs with facts. Some comments also tie it to probabilistic hardware.
Summarised by Claude from a sample of 6 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Normal Computing was founded by former members of the Google Brain Team, Palantir, and X who unlocked AI for some of the largest-scale and most mission-critical production systems, and the industry leading ML frameworks for Probabilistic and Quantum AI.
Most-linked posts?
- Supersizing Transformers: Going Beyond Rag with Extended Minds for LLMs4 people · 4 mentions · first linked 2023-11-02
- RegEx Guided LLM Generation3 people · 3 mentions · first linked 2023-08-15
- The Normal Blog – A First Demonstration of Thermodynamic Matrix Inversion2 people · 2 mentions · first linked 2023-11-10
What Hacker News says?
I’m not a fan of Extropic, but I’m seeing a lot of misconceptions here. They’re not building “a better rng”- they’re building a way to bake probabilistic models into hardware and then run inference on them using random fluctuations. Theoretically this means much faster inference for things like PGMs. See here for similar things: https://arxiv.org/abs/2108.09836 There’s a company called Normal Computing that did some…
Comment linking /posts/2023-11-09-thermodynamic-inversion/thermo-inversion.html
on “What Extropic is building”Submitted Posteriors from Normal – Approximate Bayesian inference for large models as a story · 18 points
HN discussionI wonder whether that could be fixed by using an approach like this [1] [2], to guarantee the model output meets a schema. [1] https://news.ycombinator.com/item?id=37125118 [2] https://blog.normalcomputing.ai/posts/2023-07-27-regex-guide...
Comment linking /posts/2023-07-27-regex-guided-generation/regex-guided-generation.html
on “Show HN: Chessdream – Generate realistic chess positions with AI”Submitted Supersizing Transformers: Going Beyond Rag with Extended Minds for LLMs as a story · 4 points
HN discussionRegex-Gen is implemented in all generality in the library (minus some constructs that we still have to add). JSON is merely an application. You can read https://blog.normalcomputing.ai/posts/2023-07-27-regex-guide... for a more detailed explanation of how it works. Should answer your question :)
Comment linking /posts/2023-07-27-regex-guided-generation/regex-guided-generation.html
on “Show HN: LLMs can generate valid JSON 100% of the time”Submitted Thermox: The First Thermodynamic Computing Simulator as a story · 2 points
HN discussionI'm excited to share Fuji-Web, an open-source AI web agent designed to automate various web tasks. I had this idea of using vision-LLM to build web agent project in Nov 2023 when GPT-4V was just released. I'm so proud the interesting idea has evolved into a state-of-the-art Web Agent. (You can find benchmarks in the blog post.) We started this research because we wanted to find out how far away we are from having an…
Comment linking /posts/2024-05-22-introducing-fuji-web/fuji-web.html
on “Show HN: Fuji-Web – An Open-Source AI Web Agent with SoTA Performance”Submitted Explainable Language Models: Existing and Novel Approaches as a story · 2 points
HN discussionI feel, if you use a pre-trained model to do these things without knowing the set intersection of the test and that dataset, makes it very tough to know weather inference is in the transitive closure of generated text the models were trained on or weather they really improved. There was another approach to grounding LLMs the other day from Normal Computing the other day: https://blog.normalcomputing.ai/posts/2023-09…
Comment linking /posts/2023-09-12-supersizing-transformers/supersizing-transformers.html
on “Wikidata, with 12B facts, can ground LLMs to improve their factuality”Submitted The Normal Blog – A First Demonstration of Thermodynamic Matrix Inversion as a story · 2 points
HN discussionSomething like this perhaps: https://blog.normalcomputing.ai/posts/2023-09-12-supersizing...
Comment linking /posts/2023-09-12-supersizing-transformers/supersizing-transformers.html
on “Download all of Wikipedia on your phone”Submitted Infinite Context LLMs: Going Beyond RAG with Extended Minds as a story · 1 points
HN discussionSubmitted Supersizing Transformers: Going Beyond Rag with Extended Minds for LLMs as a story · 1 points
HN discussionSubmitted RegEx Guided LLM Generation as a story · 1 points
HN discussion
