qm – Multiplayer agent harness for work
502 by tosh | 107 comments on
New best story on News: Show HN: I was tired of opening 2 tabs for every HN link, so I made a userscript
Show HN: I was tired of opening 2 tabs for every HN link, so I made a userscript
408 by twalichiewicz | 114 comments on News.
HN is great for the links people share, but a big part of the value I get comes from reading the discussion around them. I realized I was always opening the article in one tab and the comments in another, constantly switching back and forth. I figured there was probably a simpler way, so I threw together this userscript to merge the two. 1. Clicking a link from Hacker News opens the article with a side panel containing the discussion. It doesn't require your credentials, is resizable, and is easy to tweak if you want to customize it. 2. If you land on an article that has previously been shared on HN, the script finds the existing discussion and adds a button in the top-right to open the panel. Feedback welcome.
408 by twalichiewicz | 114 comments on News.
HN is great for the links people share, but a big part of the value I get comes from reading the discussion around them. I realized I was always opening the article in one tab and the comments in another, constantly switching back and forth. I figured there was probably a simpler way, so I threw together this userscript to merge the two. 1. Clicking a link from Hacker News opens the article with a side panel containing the discussion. It doesn't require your credentials, is resizable, and is easy to tweak if you want to customize it. 2. If you land on an article that has previously been shared on HN, the script finds the existing discussion and adds a button in the top-right to open the panel. Feedback welcome.
New best story on Hacker News: Show HN: I was tired of opening 2 tabs for every HN link, so I made a userscript
Show HN: I was tired of opening 2 tabs for every HN link, so I made a userscript
408 by twalichiewicz | 114 comments on
HN is great for the links people share, but a big part of the value I get comes from reading the discussion around them. I realized I was always opening the article in one tab and the comments in another, constantly switching back and forth. I figured there was probably a simpler way, so I threw together this userscript to merge the two. 1. Clicking a link from Hacker News opens the article with a side panel containing the discussion. It doesn't require your credentials, is resizable, and is easy to tweak if you want to customize it. 2. If you land on an article that has previously been shared on HN, the script finds the existing discussion and adds a button in the top-right to open the panel. Feedback welcome.
408 by twalichiewicz | 114 comments on
HN is great for the links people share, but a big part of the value I get comes from reading the discussion around them. I realized I was always opening the article in one tab and the comments in another, constantly switching back and forth. I figured there was probably a simpler way, so I threw together this userscript to merge the two. 1. Clicking a link from Hacker News opens the article with a side panel containing the discussion. It doesn't require your credentials, is resizable, and is easy to tweak if you want to customize it. 2. If you land on an article that has previously been shared on HN, the script finds the existing discussion and adds a button in the top-right to open the panel. Feedback welcome.
New best story on Hacker News: Kimi-K3 Technical Report [pdf]
Kimi-K3 Technical Report [pdf]
380 by vinhnx | 168 comments on
Related: Kimi-K3 on HuggingFace - https://ift.tt/yh82bQt
380 by vinhnx | 168 comments on
Related: Kimi-K3 on HuggingFace - https://ift.tt/yh82bQt
New best story on News: Nvidia, Microsoft, Meta warn against overregulating open-weight models
Nvidia, Microsoft, Meta warn against overregulating open-weight models
516 by louiereederson | 238 comments .
Letter: https://ift.tt/wokmIEL... [pdf] https://ift.tt/ZiBdHYS , https://ift.tt/cRMfhl0 https://ift.tt/t2npWT3... , https://ift.tt/CwME2NY
516 by louiereederson | 238 comments .
Letter: https://ift.tt/wokmIEL... [pdf] https://ift.tt/ZiBdHYS , https://ift.tt/cRMfhl0 https://ift.tt/t2npWT3... , https://ift.tt/CwME2NY
New best story on News: John C. Dvorak has died
John C. Dvorak has died
594 by coleca | 186 comments .
https://ift.tt/8SVBw17 https://ift.tt/jrMUiXO... https://ift.tt/qFa02cB...
594 by coleca | 186 comments .
https://ift.tt/8SVBw17 https://ift.tt/jrMUiXO... https://ift.tt/qFa02cB...
New best story on News: OpenAI and Hugging Face address security incident during model evaluation
OpenAI and Hugging Face address security incident during model evaluation
589 by mfiguiere | 389 comments on News.
https://ift.tt/cKmT6dN... See also Security incident disclosure – July 2026 - https://ift.tt/4Z3eF6L (9 comments)
589 by mfiguiere | 389 comments on News.
https://ift.tt/cKmT6dN... See also Security incident disclosure – July 2026 - https://ift.tt/4Z3eF6L (9 comments)
New best story on Hacker News: OpenAI and Hugging Face address security incident during model evaluation
OpenAI and Hugging Face address security incident during model evaluation
577 by mfiguiere | 380 comments on
https://ift.tt/GEiwvuK... See also Security incident disclosure – July 2026 - https://ift.tt/AE6Q70e (9 comments)
577 by mfiguiere | 380 comments on
https://ift.tt/GEiwvuK... See also Security incident disclosure – July 2026 - https://ift.tt/AE6Q70e (9 comments)
New best story on Hacker News: Thanks HN for 15 years of support and helping me find my life's work
Thanks HN for 15 years of support and helping me find my life's work
421 by nicholasjbs | 42 comments on
Tomorrow is the 15th anniversary of the first day of the Recurse Center ( https://ift.tt/I4GDFnc ) My cofounders and I did YC all the way back in the Summer of 2010, with the initial idea of building "OkCupid for jobs." That idea quickly fizzled, and we spent the better part of a year pivoting between other ideas that also failed. Finally, we made something that we wanted ourselves: a self-directed programming retreat, where people built fun projects, contributed to open source, and helped each other become better programmers. After running two small batches, we launched on HN[1] and got an incredible reception. That post on HN helped us reach beyond our personal networks and meet programmers from around the world, many of whom have since become friends. HN brought us the majority of people who came to our next few batches, and in the years since, HN has remained our #2 source of applicants (after word of mouth). Alas, pg's comment[2] on HN when we launched turned out to be prescient: Running free programming retreats isn't a billion-dollar business, but it's still a worthwhile thing to do, and has positively impacted over 3,000 people so far. And 15 years on I still wake up every day excited to keep working on it. So, thanks HN, for helping make the Recurse Center possible, and for helping me find my life's work. [1] https://ift.tt/qYPHczy [2] "This sounds like a crazy plan for a startup, I realize, but this is the right sort of crazy. In fact, the way the Hackruiters think about Hacker School is a lot like the way we initially thought about YC: if it doesn't make money, it will at least have been a benevolent thing to do."
421 by nicholasjbs | 42 comments on
Tomorrow is the 15th anniversary of the first day of the Recurse Center ( https://ift.tt/I4GDFnc ) My cofounders and I did YC all the way back in the Summer of 2010, with the initial idea of building "OkCupid for jobs." That idea quickly fizzled, and we spent the better part of a year pivoting between other ideas that also failed. Finally, we made something that we wanted ourselves: a self-directed programming retreat, where people built fun projects, contributed to open source, and helped each other become better programmers. After running two small batches, we launched on HN[1] and got an incredible reception. That post on HN helped us reach beyond our personal networks and meet programmers from around the world, many of whom have since become friends. HN brought us the majority of people who came to our next few batches, and in the years since, HN has remained our #2 source of applicants (after word of mouth). Alas, pg's comment[2] on HN when we launched turned out to be prescient: Running free programming retreats isn't a billion-dollar business, but it's still a worthwhile thing to do, and has positively impacted over 3,000 people so far. And 15 years on I still wake up every day excited to keep working on it. So, thanks HN, for helping make the Recurse Center possible, and for helping me find my life's work. [1] https://ift.tt/qYPHczy [2] "This sounds like a crazy plan for a startup, I realize, but this is the right sort of crazy. In fact, the way the Hackruiters think about Hacker School is a lot like the way we initially thought about YC: if it doesn't make money, it will at least have been a benevolent thing to do."
New best story on News: Thanks HN for 15 years of support and helping me find my life's work
Thanks HN for 15 years of support and helping me find my life's work
412 by nicholasjbs | 39 comments .
Tomorrow is the 15th anniversary of the first day of the Recurse Center ( https://ift.tt/I4GDFnc ) My cofounders and I did YC all the way back in the Summer of 2010, with the initial idea of building "OkCupid for jobs." That idea quickly fizzled, and we spent the better part of a year pivoting between other ideas that also failed. Finally, we made something that we wanted ourselves: a self-directed programming retreat, where people built fun projects, contributed to open source, and helped each other become better programmers. After running two small batches, we launched on HN[1] and got an incredible reception. That post on HN helped us reach beyond our personal networks and meet programmers from around the world, many of whom have since become friends. HN brought us the majority of people who came to our next few batches, and in the years since, HN has remained our #2 source of applicants (after word of mouth). Alas, pg's comment[2] on HN when we launched turned out to be prescient: Running free programming retreats isn't a billion-dollar business, but it's still a worthwhile thing to do, and has positively impacted over 3,000 people so far. And 15 years on I still wake up every day excited to keep working on it. So, thanks HN, for helping make the Recurse Center possible, and for helping me find my life's work. [1] https://ift.tt/qYPHczy [2] "This sounds like a crazy plan for a startup, I realize, but this is the right sort of crazy. In fact, the way the Hackruiters think about Hacker School is a lot like the way we initially thought about YC: if it doesn't make money, it will at least have been a benevolent thing to do."
412 by nicholasjbs | 39 comments .
Tomorrow is the 15th anniversary of the first day of the Recurse Center ( https://ift.tt/I4GDFnc ) My cofounders and I did YC all the way back in the Summer of 2010, with the initial idea of building "OkCupid for jobs." That idea quickly fizzled, and we spent the better part of a year pivoting between other ideas that also failed. Finally, we made something that we wanted ourselves: a self-directed programming retreat, where people built fun projects, contributed to open source, and helped each other become better programmers. After running two small batches, we launched on HN[1] and got an incredible reception. That post on HN helped us reach beyond our personal networks and meet programmers from around the world, many of whom have since become friends. HN brought us the majority of people who came to our next few batches, and in the years since, HN has remained our #2 source of applicants (after word of mouth). Alas, pg's comment[2] on HN when we launched turned out to be prescient: Running free programming retreats isn't a billion-dollar business, but it's still a worthwhile thing to do, and has positively impacted over 3,000 people so far. And 15 years on I still wake up every day excited to keep working on it. So, thanks HN, for helping make the Recurse Center possible, and for helping me find my life's work. [1] https://ift.tt/qYPHczy [2] "This sounds like a crazy plan for a startup, I realize, but this is the right sort of crazy. In fact, the way the Hackruiters think about Hacker School is a lot like the way we initially thought about YC: if it doesn't make money, it will at least have been a benevolent thing to do."
New best story on Hacker News: Apple sues OpenAI, accuses ex-employees of stealing trade secrets
Apple sues OpenAI, accuses ex-employees of stealing trade secrets
415 by stock_toaster | 201 comments on
https://ift.tt/AfbLTE0
415 by stock_toaster | 201 comments on
https://ift.tt/AfbLTE0
New best story on Hacker News: Show HN: Getting GLM 5.2 running on my slow computer
Show HN: Getting GLM 5.2 running on my slow computer
706 by vforno | 169 comments on
A few days ago I found myself trying out GLM 5.2 and was really positively impressed. The capabilities and security I was getting from this LLM are similar to those I've gotten from models like Claude or GPT, and this really surprised me. But then I thought, "I wonder how it would work on a normal computer like mine," and above all, "I wonder if it would work without going into OOM on a computer like mine." So I started working with the help of agents to test this possibility. I started converting the model to int4, understanding MTP usage, and if possible implementing DSA for long context. How it responds in int4 and whether the quality is maintained or not. Until I got to the point, on my computer with 32GB of RAM, I was able to communicate with GLM 5.2 with times that, of course, aren't high in cold start, but even then, we're talking about 0.1 tok/s, but that wasn't important to me. The important thing was the journey to reach this goal. I just wanted it to work at all costs, even slowly. So I created Colibrì, which was born from a very simple idea, to be honest, but tested in every way, where a 744B Mixture-of-Experts model activates only ~40B parameters per token—and only ~11 GB of those change from token to token (the routed experts). So: The dense part (attention, shared experts, embeddings—~17B params) stays resident in RAM at int4 (~9.9 GB); The 21,504 routed experts (75 MoE layers × 256 experts + the MTP head, ~19 MB each at int4) live on disk (~370 GB) and are streamed on demand, with a per-layer LRU cache, an optional pinned hot-store, and the OS page cache as a free L2. The engine is a single C file (c/glm.c, ~1,300 lines) plus small headers. No BLAS, no Python at runtime, no GPU.No GPU or serious hardware because I don't have that hardware so I can't test it on hardware that is more powerful than my computer.Colibrì is a one-person project, written and tested entirely on a 12-core laptop with 25 GB of RAM — the numbers above are the ceiling of what I can measure at home. Any feedback is welcome! (and if anyone wanted to participate in the project I would be delighted) Repo: https://ift.tt/fHhElJy
706 by vforno | 169 comments on
A few days ago I found myself trying out GLM 5.2 and was really positively impressed. The capabilities and security I was getting from this LLM are similar to those I've gotten from models like Claude or GPT, and this really surprised me. But then I thought, "I wonder how it would work on a normal computer like mine," and above all, "I wonder if it would work without going into OOM on a computer like mine." So I started working with the help of agents to test this possibility. I started converting the model to int4, understanding MTP usage, and if possible implementing DSA for long context. How it responds in int4 and whether the quality is maintained or not. Until I got to the point, on my computer with 32GB of RAM, I was able to communicate with GLM 5.2 with times that, of course, aren't high in cold start, but even then, we're talking about 0.1 tok/s, but that wasn't important to me. The important thing was the journey to reach this goal. I just wanted it to work at all costs, even slowly. So I created Colibrì, which was born from a very simple idea, to be honest, but tested in every way, where a 744B Mixture-of-Experts model activates only ~40B parameters per token—and only ~11 GB of those change from token to token (the routed experts). So: The dense part (attention, shared experts, embeddings—~17B params) stays resident in RAM at int4 (~9.9 GB); The 21,504 routed experts (75 MoE layers × 256 experts + the MTP head, ~19 MB each at int4) live on disk (~370 GB) and are streamed on demand, with a per-layer LRU cache, an optional pinned hot-store, and the OS page cache as a free L2. The engine is a single C file (c/glm.c, ~1,300 lines) plus small headers. No BLAS, no Python at runtime, no GPU.No GPU or serious hardware because I don't have that hardware so I can't test it on hardware that is more powerful than my computer.Colibrì is a one-person project, written and tested entirely on a 12-core laptop with 25 GB of RAM — the numbers above are the ceiling of what I can measure at home. Any feedback is welcome! (and if anyone wanted to participate in the project I would be delighted) Repo: https://ift.tt/fHhElJy
Subscribe to:
Posts (Atom)