TechniqueRLHF / Alignment8 recent entries4 Aug 2026On the Limits of Support-Preserving Alignment and Bounded FilteringarXiv:2607.18295v2 Announce Type: replace Abstract: We study whether alignment schemes that reshape a base model's output distribution, combined with bounded safety filters, can drive the probability →16 Aug 2026Qwen 3.8 27B is excellent, but it defaults to wildly overthinking thingsFriday's big release was Qwen 3.8 27B, an Apache 2 licensed 27B parameter vision-capable LLM from Alibaba's Qwen research lab. I've been looking forward to this one: 27B is an excellent size for runni
TechniqueRAG8 recent entries15 Aug 2026Agentic harness for small modelsHello; I'm a semi-beginner at local AI. I've been experimenting with this tech for a while, and I still haven't found a proper harness that fits my models, hardware, and needs. My use case is pretty s→7 Sept 2026RefactorPlatform: An Open-Source Harness for Controlled Evaluation of Repository-Scale Refactoring AgentsarXiv:2609.04898v1 Announce Type: cross Abstract: Repository-scale refactoring requires coding agents to propagate a single change across many interdependent files without altering program behavior, y→9 Sept 2026Wiki Lint Report — 2026-09-09Automated lint: 12 errors, 53990 warnings, 3 info→13 Sept 2026Wiki Lint Report — 2026-09-13Automated lint: 12 errors, 54839 warnings, 3 info→20 Sept 2026Wiki Lint Report — 2026-09-20Automated lint: 14 errors, 54840 warnings, 3 info→27 Sept 2026Wiki Lint Report — 2026-09-27Automated lint: 17 errors, 59669 warnings, 3 info→4 Oct 2026Wiki Lint Report — 2026-10-04Automated lint: 19 errors, 64915 warnings, 3 info→12 Oct 2026Qwen3.8-27B vs Frontier vs OthersHere're results from my final test of coding, coding (panda/numpy), data analysis, data-driven decision-making, and local RAG domains I interact with most often. This test, which is refined from a 500
TechniqueAgents8 recent entries5 Oct 2026Own your personal agent, folks. Just built my own little personal agent with @pidotdev's Pi Durable. I wanted to see how quickly I could bui…Own your personal agent, folks. Just built my own little personal agent with @pidotdev's Pi Durable. I wanted to see how quickly I could build an OpenAI dots clone with open-source tools. I am amazed →6 Oct 2026We just raised $7.7M to kill every subscription for agents. Introducing http://monid.ai, the openrouter for agent tools. Your agents can now…We just raised $7.7M to kill every subscription for agents. Introducing http://monid.ai, the openrouter for agent tools. Your agents can now: > discover, run, and pay for tools at runtime > use 2,500 →6 Oct 2026i was told we can't fundraise and grow at the same time. it turned out we grew monthly usage 400% during fundraising. we have never run this…i was told we can't fundraise and grow at the same time. it turned out we grew monthly usage 400% during fundraising. we have never run this fast. while taking 60 investor calls a week, we shipped 20 →6 Oct 2026Claude Opus 5.5 leads at 63.31 and 4.99 per task. GPT 6 Astra is second at 56.25 and 11.61. Sonnet 5.5 follows at 53.14 and $2.82. At the …Claude Opus 5.5 leads at 63.31 and 4.99 per task. GPT 6 Astra is second at 56.25 and 11.61. Sonnet 5.5 follows at 53.14 and 2.82. At the other end, DeepSeek V4.1 Flash scores 36.91 at 0.26 and Ling 3.→7 Oct 2026Btw, we're hiring. We're hiring the best engineers in the world. You're ambitious. Want to make something different. You learn and build fas…Btw, we're hiring. We're hiring the best engineers in the world. You're ambitious. Want to make something different. You learn and build fast. Our company is growing fast, so you need to grow even fas→8 Oct 2026OpenDocRouter is basically OpenRouter for document OCR You can compare and use all your favorite OCR models under a single API / billing int…OpenDocRouter is basically OpenRouter for document OCR You can compare and use all your favorite OCR models under a single API / billing interface. They range from some of the lightweight/OSS models (→9 Oct 2026rate the terminal setupik ik 'another harness D:'. so... i spent the last 6 months of my life building and ironing out every single detail and bug through almost every route with extensive testing across 3 machines, daily, →9 Oct 2026My 'Anthropic/ChatGPT in a box' is... now open source and publicly hosted on Github.It's been a long time coming, and I should have done this in the first place. But LumaBrowser is now open source. Since I first released it a year ago, it's had several thousand installs and an absolu
TechniqueFine-tuning8 recent entries20 Sept 2026Wiki Lint Report — 2026-09-20Automated lint: 14 errors, 54840 warnings, 3 info→27 Sept 2026Wiki Lint Report — 2026-09-27Automated lint: 17 errors, 59669 warnings, 3 info→2 Oct 2026Qwen3.8-27B-Humanlike-Chat 2.0: texts like a human, now with tool calls and better instruction followingLast month I posted a Qwen3.8-27B LoRA that makes it talk like a person instead of an assistant. It got a lot more attention than I expected: 700+ upvotes, 248 comments and 44k downloads since. I read→3 Oct 2026ai slop game based on batlle video in real time with only open modelCheck out Chimera Arena (https://chimeraarena.com/static/showcase/index.html?lang=en)! Here’s my slightly scrappy AI-generated teaser for Chimera Arena, my card game combining AI-generated video battl→4 Oct 2026Wiki Lint Report — 2026-10-04Automated lint: 19 errors, 64915 warnings, 3 info→7 Oct 2026Ruach Studio: a whole song studio around YuE2 on your own GPU. Score first, LoRA training, stems,remaster, DAW export.We spent the last weeks building a studio around YuE2, the open song model by m-a-p, and today it reaches its first release candidate. Write the style and the lyrics, and it composes, sings and render→8 Oct 2026Comparing Qwen3.8-27B fine-tunes and baselining vs. frontierTL;DR: For my use case, mradermacher/Signal-3.8-27B-Terse-Coder.i1-Q4_K_M proved out. Ymmv based on your domain-specific tests. Time taken to solve problems compared to frontier models is massive; esp→12 Oct 2026Qwen3.8-27B vs Frontier vs OthersHere're results from my final test of coding, coding (panda/numpy), data analysis, data-driven decision-making, and local RAG domains I interact with most often. This test, which is refined from a 500
TechniqueMultimodal8 recent entries9 Sept 2026Wiki Lint Report — 2026-09-09Automated lint: 12 errors, 53990 warnings, 3 info→11 Sept 2026Fugu Max is now live on @OpenRouter! 🐡 Our learned multi-agent orchestration engine routes tasks across open-weights and specialized models…Fugu Max is now live on @OpenRouter! 🐡 Our learned multi-agent orchestration engine routes tasks across open-weights and specialized models for Pareto-frontier efficiency. • 2/6 per 1M tokens • Multim→13 Sept 2026Wiki Lint Report — 2026-09-13Automated lint: 12 errors, 54839 warnings, 3 info→18 Sept 20266 billion tokens for $65 Is it as good as Astra and Fable at hard stuff? No of course not. But you don't always need a genius and can just c…6 billion tokens for $65 Is it as good as Astra and Fable at hard stuff? No of course not. But you don't always need a genius and can just call one for advice when you do. Extrapolate where this is go→20 Sept 2026Wiki Lint Report — 2026-09-20Automated lint: 14 errors, 54840 warnings, 3 info→27 Sept 2026Wiki Lint Report — 2026-09-27Automated lint: 17 errors, 59669 warnings, 3 info→4 Oct 2026Wiki Lint Report — 2026-10-04Automated lint: 19 errors, 64915 warnings, 3 info→8 Oct 2026Jevjitsu: Or, How We Tried Generalized Classifiers on EverythingClassification is one of the oldest problems in machine learning. Unlike modern decoding transformers, most classification models were non-autoregressive by nature. Before Jev, the common approaches w
TechniqueSafety8 recent entries9 Sept 2026Wiki Lint Report — 2026-09-09Automated lint: 12 errors, 53990 warnings, 3 info→13 Sept 2026Wiki Lint Report — 2026-09-13Automated lint: 12 errors, 54839 warnings, 3 info→20 Sept 2026Wiki Lint Report — 2026-09-20Automated lint: 14 errors, 54840 warnings, 3 info→26 Sept 2026This is one of the strongest use cases I've seen for System One models like Jev. Pay close attention if you build custom harnesses. Jev is n…This is one of the strongest use cases I've seen for System One models like Jev. Pay close attention if you build custom harnesses. Jev is now available as a packaged router. What is it? You send the →27 Sept 2026Wiki Lint Report — 2026-09-27Automated lint: 17 errors, 59669 warnings, 3 info→3 Oct 2026.@OpenRouter co-founder Alex Atallah, in his first podcast since Stripe acquired the company, joins @Replit co-founder Amjad Masad and a16z'….@OpenRouter co-founder Alex Atallah, in his first podcast since Stripe acquired the company, joins @Replit co-founder Amjad Masad and a16z's Erik Torenberg on why the future of AI is independence and→4 Oct 2026Wiki Lint Report — 2026-10-04Automated lint: 19 errors, 64915 warnings, 3 info→8 Oct 2026Jevjitsu: Or, How We Tried Generalized Classifiers on EverythingClassification is one of the oldest problems in machine learning. Unlike modern decoding transformers, most classification models were non-autoregressive by nature. Before Jev, the common approaches w