AI Intelligence Archive

As the Machine Learning Engineering Manager at an AI-powered SaaS company, I get a front-row seat to...

Reinforcement learning (RL) for learning path recommendation (LPR) faces two coupled obstacles. First, the policy must commit to a sequence of L concepts without intermediate feedback, producing a combinatorial search space that grows supe…

"Task completed successfully." That's the most common lie in agent systems. Usually the agent isn't...

Touché 2026 extends causality extraction to counter-causal claims: news sentences whose surface form appears causal but whose meaning denies the causation, as in "It is falsely believed that X caused Y." A system that relies on surface cue…

I’ve been a backend dev for about four years, and lately, I feel like I'm attending the funeral of my...

Large-scale pre-training has improved the generalization of neural operators across diverse PDEs. However, existing PDE foundation models still struggle with heterogeneous dynamics, where shared representations may cause knowledge interfer…

I work on proxy infrastructure, which means I read a lot of proxy errors. And I kept noticing the...

Solving severely ill-posed imaging inverse problems requires recovering image structures that are unobservable or weakly constrained by the measurements. Diffusion models provide expressive learned priors for inferring such missing informa…

We release PaMIR (Public Arrival-ordered Measurement for Inference in Risk), an open benchmark for credit-default prediction when labels are scarce and arrive late. The field's reference benchmark studies use eight datasets each, only two …

Scalable nonlinear causal discovery requires methods that combine flexible mechanism estimators with efficient search over large graph spaces. Several algorithmic families have been proposed to address this challenge, yet their accuracy-ru…

Coding agents write their own session transcripts to disk. Where Claude Code, Codex and OpenCode keep those files, and how to find the session that belongs to your task.

We present a first evaluation of machine learning applied to patient clinical and demographic data gathered in two different countries for the purpose of tuberculosis (TB) screening to identify people who would benefit from expensive molec…

A model trained on "The capital of X is Y" may produce "Y" after "The capital of X is" but fail after "The capital of X:". We call these different ways of eliciting the same fact request forms. To separate learning a fact from retrieving i…

I have three daughters. The oldest is four, the twins are two. None of them can read yet, and every...

There was a point while building Security Lens when I realized I was fixing problems faster than I...

Large language models (LLMs) are increasingly deployed in public health, finance, and governance, requiring both accuracy and societal value alignment. Despite recent advances, LLMs often perpetuate or amplify bias embedded in their traini…

We present WAMpy, a Python framework optimized for synthesizing Prolog programs. Unlike general-purpose Prolog systems, WAMpy targets workloads that repeatedly generate and evaluate small candidate programs. WAMpy compiles Prolog clauses i…

GPU kernels generated by large language model (LLM) agents can remain less efficient than expert implementations, but runtime alone does not reveal how the gap relates to design discovery and implementation. We introduce D2K-Bench, a diagn…

We present the Neuro-Physical Inverter (NPI), a modular, uncertainty-aware framework for geophysical inversion that couples ensemble-based conditioning with constrained residual learning, demonstrated in the 1D magnetotelluric (MT) setting…

Diffusion models can synthesise contrast-enhanced CT (CECT) from non-contrast CT (NCCT), avoiding contrast administration and its environmental and patient-access costs. However, visually realistic images are not necessarily anatomically c…

Long-horizon LLM agents are typically trained with sparse outcome rewards, making trajectory-level objectives too coarse to distinguish the contribution of individual decisions. Step-level credit assignment provides finer-grained supervisi…

This paper studies the complexity of convex optimization using lazy second-order oracles (Doikov, Chayti, and Jaggi, ICML 2023), where an algorithm queries gradients every iteration and Hessians once per $m$ iterations. Under this setting,…

Hybrid quantum classical neural networks integrate parameterized quantum circuits (PQCs) with established deep learning architectures, but their performance depends strongly on the choice of quantum circuit architecture, a choice that rema…

TL;DR A partnership for security: Our new Sandbox Mixin Kit pairs Docker's isolated...

How to deploy a private, local AI gateway for engineering teams using LiteLLM, Caddy, Langfuse, and Open WebUI with multi-tier quotas.

For the Sanity Challenge, I decided to tackle Path One: Ship an agent that queries real content. As...

ghswarm treats GitHub labels, Issue bodies and checkboxes as the state machine for orchestrating Claude Code / Codex / Cursor headlessly. No webhook, no FastAPI, no database — and it resumes cleanly after the process dies.

Every tool you connect to an agent is two things at once: tokens the model reads on every run, and an...

I started my degree just 1-2 months ago, which means I will be graduating in 2030. Given how rapidly...

On 2026-08-01 I computed how long an AI model lives, from the day it shows up to the day its vendor...

Helping students find traceable sources, understand their data, and develop research drafts they can review and defend. Powered by an open-source Node.js agent harness, Google Gemini, and SerpApi.

One month ago, I saw someone's hobby OS and thought: "I want to build one too." One month later,...

A real failure story from building EscrowLens, a passkey-native escrow protocol on Monad for the...

This is a submission for the Kaggle Benchmarking Challenge Most incidents start with a graph. Not a...

I have been doing open-source for 10 years now. From side projects to small and popular libraries...

I was looking for a relaxed, developer-first DevOps event close to Finland, and DevOpsDays Vilnius...

This is a solo submission by LOI CHIANG HAO for the Sanity Challenge, Path One: Ship an Agent That...

Until this week, CodeSage ran its two models exactly as they ship on Hugging Face: Jina's code...

This is a submission for the Sanity Challenge, Path One: Ship an Agent That Queries Real Content. ...

On Wednesday I posted 187 live prompts, 27 bugs about testing CORTEX, my local AI agent, against a...