AI Intelligence Archive

Audit of the 78-row SWE-bench Verified board on Hugging Face: 0 verified, 51 open PRs, 9 rows above swebench.com's best, one repo spanning 23 points.

Five days after I published the docs for my app, 36 of its 37 pages described code that had changed...

Why do New Caledonian crows bend wire into hooks with zero training while a trillion-parameter LLM flies straight into the same production bug zapper? How Anthony Zador's Genomic Bottleneck explains what 'instinct' really is in AI coding a…

Yonto Player is a media player for your own library: your Jellyfin, Plex or Emby server, or the files...

Organic discovery is becoming a less dependable single source of website visits in the data examined...

git commit -a holds .git/index.lock while your pre-commit hook runs. SIGKILL the agent then and the lock stays. Plus the locks every worktree shares.

Disclosure: I built LaunchVet, the tool I use in this article. The free core is MIT licensed. Every...

Walmart’s new approach to AI agents points to a consequential change in how automated assistants may...

We've all been there. You join a new team, open the repository, and find a beautifully written...

We're excited to announce our newest challenge with Presage! The Presage SmartSpectra SDK Challenge...

Google has expanded AI Mode in Search to surface direct booking and quote links for home services in...

Here are two sentences that mean exactly the same thing: The government announced a new policy on...

Garuda is an open-source coding agent for the terminal that I build to learn how coding agents work....

Microsoft found only 14% of review comments flagged defects. When an agent writes the diff, review is the one moment a human learns the code.

Models got dramatically better over the past year. But they got better for everyone. In that time, our team went from 5.3 to 13.7 PRs per person per month, and I personally went from 4.8 to 51.3. Did you improve that much too? If not, here…

Every AI trend comes with a prompt. A sherwani portrait, 3D balloon letters, a pet dancing next to a...

Originally published on mihai.ltd. How many times has an AI coding agent written a feature that...

Toda demo de agente termina em 'pronto' sem prova. Gate cego faz o 'pronto' valer alguma coisa: o artefato é validado fora do alcance de quem construiu.

Journey Before Destination I got into software engineering because I enjoyed it. I enjoyed the...

AI-DLC, bolts, and autonomous agents promise faster software delivery. But who owns the architecture when AI starts making the decisions?

Originally published by the AI Visibility Census at...

A runnable evaluation example separating wrong answers, memory limits, timeouts and setup failures, with all 27 attempt records retained.

Serving a long-context model is a memory problem disguised as a compute problem. Offloading the cache to NVMe and recomputing the prefill are both defensible, and they fail in completely different way

What If Your AI Coding Agent Deletes Your Project? Building a Zero-Trust Sandbox with...

Moonshot AI（@Kimi_Moonshot）在 HuggingFace 开放 Kimi K3.1 Pro 权重（https://huggingface.co/moonshotai/Kimi-K3.1-Pro）。

Ask a language model for an AWS architecture and you get a confident answer. Sometimes it includes a...

Have you ever had a dev server running, and then asked your coding agent to test something in that...

Researchers use large language models as agents in social simulations, where agents are prompted with observations and background information intended to shape their actions. Yet LLM agents have been shown to substantially change their beh…

Reconstructing full flow fields from sparse sensors requires prior assumptions about unmeasured regions and physical variables. Yet it remains unclear when learned models outperform classical estimators and which aspects of their predictio…

Self-Driving Chemical Laboratories (SDCLs) are moving chemical experimentation toward embodied automation, where machines must perceive laboratory scenes, reason about experimental states, and act under physical, procedural, and safety con…

AI agents are playing an increasingly important role in physics research. However, existing physics benchmarks only evaluate fixed questions, isolated sub-tasks, or paper reproduction with given solutions, and rarely check research quality…

Humanoid robots have the potential to automate a wide range of physical tasks. Progress toward generalist humanoid autonomy, however, remains constrained by a data bottleneck: mapping pixels and free-form language to low-level joint comman…

The introduction of generalist robot policies has enabled a single, large model to learn to perform a variety of tasks from a vast pre-training corpus. While these models often exhibit powerful zero-shot generalization capabilities, they t…

Knowing that a pedestrian is occluded is useful to a robot even when the detector cannot recover the hidden pixels, but occlusion estimators are normally trained on occlusion annotations that deployment settings rarely have. We show that o…

Automation is often justified as relief from dull, dirty, and dangerous work. I examine how these conditions relate to occupational AI exposure using 21 O*NET items across 754 US occupations. Higher exposure is associated with lower condit…

The best classical optimizer for a continuous black-box problem depends on its budget and landscape. Existing routers decide from the dimension and budget, past improvement or features fitted to one benchmark, and few test an optimizer's i…

LLM agents use context compaction to manage growing interaction histories, but summarizing long histories with large models can introduce substantial delays. Compaction therefore requires balancing information retention against latency. We…

Error correcting codes are a fundamental component of communication systems, enabling reliable data transmission over noisy channels. Recently, neural decoders have demonstrated notable performance gains over classical iterative algorithms…

A strategy built from a public ledger of losses: four lanes ranked by measured EV, the gate that fences risk capital, and the arithmetic that decides whether $10/day is reachable.

In this work, we evaluate a simple inference-time scaling approach based on \textit{turn-level best-of-N sampling} (\method) for large language model (LLM) agents on complex, long-horizon tasks. We instantiate this using a \textit{generato…