research

Pay close attention to custom harnesses. This is a super interesting paper showing the impact of training inside the harness. Qwen3-14B goes from 22.2% to 54.8% on Spider 2.0-SQLite when it is trained

Interesting paper from Sakana AI on memory for recurrent models. Recurrent models are good at state tracking, but they usually keep everything in one hidden vector, so short-term computation and long-

Announcing ARC Prize Research Summit 2026 Research Speaker Kevin Murphy @sirbayes is a research scientist at Google DeepMind and author of the “Probabilistic Machine Learning” textbooks, developing AI

I find it odd watching armchair critics rush to nitpick @MistralAI’s plans or obsess over whether they’re #1 on every leaderboard. Having a team with this level of conviction building frontier models

Ben Affleck reveals he writes Python, understands convolutional neural networks, worked extensively with GPUs, and used his celebrity status to get private looks at Google and OpenAI’s video models “I

Spontaneous morning chat with Juergen Schmidhuber and Jake Newfield philosophizing about AI. Sry my mic somehow disconnected but Juergen had the more interesting points anyways! Our CEO @Jake_Newfield

@abuchanlife - training frontier models is expensive - distilling a frontier model is cheap - this market force alone tells us that frontier foundation models will be free/open. - there are others. ht

The AI research community released a set of notable papers between September 28 and October 4, highlighted in a weekly roundup. Key contributions include JAZ, CASD, Agensh, Jev‑Mem, AutoGym, Taste‑Ben

Banger paper from CMU on harness learning. (bookmark it) Also, pay attention to this important new AI engineering skill of improving agents by editing their harness code instead of their weights. Seei

Language is not fundamental to human consciousness. In terms of evolution, it arrived only recently on our timeline. Consciousness evolved over millions of years through sensory awareness, emotion, an

Side note. The idea that current models are conscious is absurd and downright offensive. Do not confuse competent information processing and sentience, they are wholly unrelated. You can be entirely i

Cool eval. Simply ask an LLM “Land or Water?” and give it a latitude and longitude coordinate as text. Ask 16,200 times, plot as image. The models know. From compressing the internet. results for all

Ten years ago, AlphaGo’s Move 37 shocked the world. It wasn’t intuition alone that produced it. AlphaGo could search possible futures, test its instincts and reason about what would happen next. In a

We'll be spending a lot more time trying to understand the outputs of language models. A few thoughts, tips & tricks: Writing. Something I've had success with: Ask your LLM to explain something in ASD

During a Swyx podcast interview in October 2026, an OpenAI engineer revealed that the company’s Decision API is built on the GPT‑6 Luna model, augmented with a dedicated decision head. This configurat

Congrats to @hardmaru @risi1979 @yujin_tang and Risto! Neuroevolution is where I got started in AI, and neuroevolution is still a playground for original ideas with neural networks. This book is worth

arXiv:2609.35177v1 Announce Type: new Abstract: Estimating integrals of black-box, high-dimensional functions, from expectations and kernel mean embeddings to the softmax kernel in self-attention, is

GPT-6 Sol from @OpenAI on ARC-AGI (Verified): - ARC-AGI-3: 4.6%, 5.6K (standard harness), 23.0%, 8.7K (provider adapter harness) - ARC-AGI-2: 89.6%, 0.44/task - ARC-AGI-1: 95.5%, 0.14/task Sol reaches

Neuroevolution is a special field within AI/ML -- owe my career to it, and is underappreciated relative to its downstream impact (similar to ALIFE) -- helped birth ideas like hypernetworks, quality di

A little Ember-1 tl;dr. Seems like a great model! (Was recently asked on a podcast, given $ xx million, what's the best way to develop a frontier LLM today? My recommendation was: start with an existi

I've been suspicious of the philosophers hired by Anthropic who specialize in AI ethics for similar reasons that I'm skeptical of those who went into 'bioethics': they form a closed subculture that de

Reasoning from scratch, round number 5! This time, talking about log-probability scoring (also a great fundamental concept for loss functions like cross-entropy in pre-training and distillation) and s

Yann LeCun, Executive Chairman of AMI Labs, explains why LLMs are mostly retrieving human knowledge rather than thinking for themselves: LeCun starts with why so many people misread what these systems

Automated AI theorem proving has moved the frontier: The holy grail is not an AI that can produce an endless pile of true theorems. It’s an AI that can discover mathematics, and build each discovery i

Banger paper introducing Jev-as-a-Judge. The overall finding is that you want to use a cheap judge for most of your evals and send only the uncertain calls to a frontier model. This paper measures how

Impressive paper from Salesforce. It discusses the importance of good verifiers for RL environments. Only 35.8% of the environments in the cleanest public RL collection for terminal agents passed Sale

LLM next-token prediction is already a probabilistic classifier, so an open LLM can serve a Jev-like interface: discrete options in, fast probabilities out. Our own @ekzhang1 made it better at the job

arXiv:2609.26270v1 Announce Type: cross Abstract: We introduce the category mathbf{SSS} of sample-smooth spaces over a mixed site. The test objects are the products Omega_n := R^n imes Omega of a Cart

Finally, a company brain that actually works! Codos starts with the people doing the work. It interviews employees to find which tasks can be automated, then deploys the automations and keeps learning

One hope for Tinker is that it can handle all the ML infra for posttraining for (approximately) everyone. So, curious, what prevents you from using Tinker for all your posttraining needs? Speed? Cost?

From @WSJopinion: The Hugging Face hack wasn’t what it was cracked up to be. Forget the “hive mind” of AI agents “going rogue.” They did what humans programmed them to do, writes Brian Gross. https://

'People built the test, removed restraints, defined the objective, left a route open and decided not to stop what was happening. Calling the result 'rogue AI' does more than sensationalize it. It allo

arXiv:2609.17578v1 Announce Type: cross Abstract: Let a_1,ldots,a_m be distinct elements of a field K, and let Z(X)=prod_{i=1}^m (X-a_i). We study the arithmetic complexity of computing the unique nor

Excited to announce that the Korean translation of 'Deep Learning with Python (3rd Edition)' by @fchollet and @mattdangerw is published! 🚀 Updated for Keras 3, TensorFlow, PyTorch, JAX, and Generative

Today everyone is talking about Recursive Self-Improvement (RSI). In 1987, when compute was 100,000,000 x more expensive, I published the 1st concrete RSI algorithms. Now compute is cheap, and RSI is

The Anthropic grift machine! FED BY FEAR THEATER. Take a moment and sit with @kevinnbass chart here. You do not need to be a forensic accountant to understand this money laundering. If this is not pro

arXiv:2609.13100v1 Announce Type: cross Abstract: When providing forecasted probabilities with a predictive model, the ideal model offers perfect calibration: the true probability of the outcome (i.e.

Banger paper from Meta. This work shows that byte-level models start out behind token models and then pass them as compute grows. They show this for distilled 1B models trained on up to 1 trillion byt

One thing will defeat the AI Doomers easily: Time. Little by little things will change and AI will diffuse into everyday people's lives and they'll realize it wasn't the scary bedtime story Harry Pott

2019: “GPT2 is groundbreaking in two ways. One is its size, says Dario Amodei, OpenAI’s research director. The models “were 12 times bigger, and the dataset was 15 times bigger and much broader” than

【AIの難問は、生命の難問へ：英国王立協会が紐解く「世界モデル」と知能の未来】 近年のAIの急速な発展により、AGI、すなわち人間並みの知能はすでに実現したという意見も聞かれるようになりました。果たしてそうなのでしょうか。 鍵となるのが「世界モデル（World Model）」の概念です。 世界モデルとは、生き物やAIが外の世界を内部に写し取り、次に何が起きるかを予測して行動するための土台となるもの

I must be among an extremely small group of people (n=1?) that have both 1) trained a frontier LLM and 2) designed and synthesized custom viruses in a lab with my own two hands. And I think that the t

Investors in Anthropic founded Open Philanthropy, the leading org promoting AI Doom They have a stake in regulation on AI that excludes competitors OP is the major funder of Tarbell Tarbell Fellows pe

Looped transformers are a popular architecture topic right now. This new technical report extends the loop across tokens. Recurrent Looped Transformer (RLT) makes the decoder recurrent over every toke

I love this and really hope we can come together as an industry and make it happen. We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan

Fugu Max pushes the Pareto Frontier for price-performance: • 2 input / 6 output per 1M tokens. • Performance within striking distance of elite models at 2x to 6x lower cost. • Best overall score on 6