AI news · analysis · research
Latest edition: Oct 22, 2026
News
Oct 22, 2026 · 1 min read
Articles
Oct 20, 2026 · 2 min read
Articles
Oct 18, 2026 · 2 min read
Research
Oct 17, 2026 · 1 min read
The latest AI papers, straight from the preprint server.
Articles
Why language models need retrieval, how a RAG pipeline is put together, and where it tends to fail.
Oct 20, 2026 · 2 min read
Articles
Tokens, embeddings, attention and next-token prediction: the ideas behind every modern language model.
Oct 18, 2026 · 2 min read
Articles
Agents are language models that work in a loop with tools. Here's what that means and how to build one responsibly.
Oct 15, 2026 · 1 min read
Research
The paper that introduced the transformer, what it claimed, and why it mattered so much.
Oct 17, 2026 · 1 min read
Research
Two influential papers on how model size, data and compute trade off, and how the second one changed practice.
Oct 14, 2026 · 1 min read
MIT NewsLabOct 22, 2026
This design software reimagines everyday objects as self-aware devices
The VergePressOct 22, 2026
Meta’s AI could still scrape news sites, legal experts say
The VergePressOct 22, 2026
Meta turns Muse sharing back on
WiredPressOct 22, 2026
The Social Society Helping Women Join the AI Boom
The VergePressOct 22, 2026
Sure, let’s smoosh SpaceX and Tesla together, who cares
TechCrunchPressOct 22, 2026
1Password launches a way for AI agents to log in on your behalf without seeing your password
OpenAILabOct 22, 2026
Build more responsive apps with the Decisions API
TechCrunchPressOct 22, 2026
Canva releases its first image generation model
TechCrunchPressOct 22, 2026
OpenAI makes its Jev competitor widely available as decision model competition heats up
The VergePressOct 22, 2026
Discord now formats Markdown previews right
LongZero: Enhancing Long-Context Reasoning in LLMs from Zero Data
Reinforcement learning (RL) is an effective way to strengthen the long-context reasoning of large language models (LLMs), but it needs training data that are long, diverse, challenging, and verifiable. Human annotation of such data is expensive, naturally occurring data are too s…
Think Harder Where It Matters: Item-Level Adaptive Recursion for Sequential Recommendation
Existing sequential recommenders usually allocate a fixed computational budget to every item in a user's interaction history, regardless of the item's contribution to the final prediction. We challenge this de facto design with two diagnostic findings. (1) Non-uniform computation…
PhysiCalWorld: In-Context Physics Calibration for World Models
Existing world models typically perceive a scene through current visual frame or a narrow history window. However, these observations may not reveal the physics properties that determine transition dynamics, such as an object's mass distribution, joint damping, or surface frictio…
A Quasipolynomial Algorithm for $\ell_\infty$ Contractions
We give an algorithm for computing an $\eps$-approximate fixed point of any $d$-dimensional nonexpansion with respect to the $\ell_\infty$ norm in time $\log(1/\eps)^{O(\log^2 d)}$. At its core, our method uses a boosting argument to geometrically shrink a search region. Among ot…
Self-normalized Bernstein Inequalities for Regenerative Processes and Applications to Learning Without Independence
We prove a self-normalized Bernstein inequality for bounded functions of a regenerative process with independent, identically distributed marked cycles of finite mean length. The bound uses observed cycle lengths and retains explicit terms for truncation and the cycle ending afte…
The role of locality and weight sharing in convolutional neural networks
The success of convolutional neural networks is often attributed to locality and weight sharing. However, how these two properties separately shape learning, and how their effects depend on the spatial correlations of natural images, remain poorly understood. We address these que…
