661
Why don't machine learning research agents overfit?
New research indicates that AI agents learn compressible models of data, which don’t have enough space to enable memorization.
Topics /
AI & Machine Learning
The most popular blog posts about ai & machine learning on Hacker News. 763 posts.
662
Previewing the Model Hardware Standard
Anthropic is opening a research preview of the Model Hardware Standard (MHS), a shared specification for AI agents to safely operate physical devices, to a first group of scientific research labs and advanced manufacturers.
663
Meta's Omnilingual MT for 1,600 Languages
Advances made through No Language Left Behind (NLLB) have demonstrated that high-quality machine translation (MT) scale to 200 languages. Later Large...
664
Attack of the Vibe Coded Photoshop Clones
Over the last several months I have been tracking AI guys trying to create vibe coded clones of popular apps, especially Adobe’s apps which have always been the most pirated apps since the 90s. There are now about a dozen vibe coded attempts to clone Photoshop and some other Adobe apps, not incl
665
Keeping 20k GPUs healthy
How we do active and passive monitoring on hyperscalers and neoclouds.
666
Thomson Reuters Launches Its Own Frontier Model
Thomson, the company's proprietary LLM, was trained and is run at a fraction of the cost of comparable frontier models and remains fully owned and controlled by Thomson Reuters
667
Advancing AI Benchmarking with Game Arena
We’re expanding Game Arena with Poker and Werewolf, while Gemini 3 Pro and Flash top our chess leaderboard.
668
Why we write our own C and C++ inference engines
LocalAI is the open source AI engine. Run any model, LLMs, vision, voice, image and video, on any hardware. No GPU required.
669
Wiki Builder: Skill to Build LLM Knowledge Bases
Wiki Builder is an open-source Claude Code plugin that turns the LLM-knowledge-base workflow into a one-command setup. Scaffold a configurable wiki, point Claude at sources, and let the agent compile durable pages, indexes, and concept maps for you.
670
671
Everyone is building LLM routers, we deprecated ours
We don't believe in model routing anymore. For most use cases, sticking to a single battle-tested model is the best thing you can do.
672
GPT 5.6 Cyber
Meet GPT-5.6-Cyber, OpenAI’s cybersecurity-specific model available through Daybreak Red for authorized vulnerability research, exploit validation, and security testing.
673
Using proxies to hide secrets from Claude Code
How to use HTTP proxies to prevent credential exposure in Claude Code sandboxes.
674
SubQ 1.1 Small
Subquadratic is a frontier AI research and infrastructure company building a new class of LLMs.
675
Expensively Quadratic: The LLM Agent Cost Curve
Cache reads are quadratic and dominate your long agentic conversations.
676
Minions – Stripe's Coding Agents Part 2
Minions are Stripe’s homegrown coding agents, responsible for more than a thousand pull requests merged each week. Though humans review the code, minions write it from start to finish. Learn how they work, and how we built them.
677
What Happens When the Cost of Intelligence Drops 100x
A given level of LLM intelligence costs about a thirtieth of what it did a year ago, and the rate of decline is accelerating. Using Artificial Analysis's…
678
Running local LLMs offline on a ten-hour flight
Update, 29 April 2026 This post was picked up on Hacker News and crossed 100 comments. I love that this sparks emotions. As always, there is a fair mix of reactions, which is exactly what you would expect from HN. Scroll to the bottom for extra responses to some of the comments. … Cloud Next was intense. I wrote the…
679
Changes to Sourcehut's terms of service regarding LLMs
sourcehut is a network of useful open source tools for software project maintainers and collaborators, including git repos, bug tracking, continuous integration, and mailing lists.
680
Making LLM Training Faster with Unsloth and NVIDIA
Learn how NVIDIA helped Unsloth to make fine-tuning AI models 20% faster with explanations and diagrams.
681
682
Backprop Alternative: Augmented Lagrangian Predictive Coding
A local alternative to backpropagation. PC-ALM trains residual MLPs up to 1000 layers, nearly matching backprop's performance despite using only layer-local dynamics. PC-ALM equips each layer with a feedback control dynamical system that distributes and propagates supervision credit throughout a network.
683
684
Agora-1: The Multi-Agent World Model
Agora-1, a multi-agent world model, enables multiple participants—human or AI—to share and interact within the same world simulation in real-time.
685
Our approach to age prediction
ChatGPT is rolling out age prediction to estimate if accounts are under or over 18, applying safeguards for teens and refining accuracy over time.
686
687
Gemma-4-26B-A4B with 39-69% less thinking at same accuracy
This is the fifth model in the ThinkingCap series. We reduced how much Google's gemma-4-26B-A4B-it thinks, without compromising the quality of its answers.
688
There is no posttraining data commons
The LLM revolution ran on a pretraining commons that humanity built for other reasons. Posttraining consumes tasks, not documents, and no such commons exists. If the open models community wants...
689
The assistant axis: situating and stabilizing the character of LLMs
Who is the Assistant? We investigate the character that most modern language models inhabit when interacting with users.
690
Building voice agents with Nvidia open models
Build an ultra-low-latency voice agent with NVIDIA open models. Learn how Nemotron Speech ASR achieves sub-25ms transcription, how Nemotron 3 Nano LLM and Magpie TTS work together, and how to optimize architecture for real-time voice AI deployment.