Full-text search across 12643 articles. Combine with topic and date filters; results sorted by relevance.
Developer emnlmn released Snap, an open-weight, MIT-licensed local decision engine positioned as a drop-in replacement for the TypeSafe System One API, running the 2B-parameter snap1-2b model in a single forward pass to …
Researchers introduced Mamushi, a non-parametric framework for distributional unlearning that ranks forget examples using a probabilistic classifier whose Bayes-optimal logit equals the forget-to-retain log-density ratio…
A paper published in October 2026 by Kirill Brilliantov, Alejandro Hernández-Cano and Emmanuel Abbé finds that open-source state-of-the-art MLE agent harnesses provide no advantage over a single session of a minimal-harn…
Functional ultrasound imaging (fUSI) can deliver up to 10 times higher linear resolution and 1000 times more voxels than functional MRI while requiring no magnet, according to a deep dive by Patrick Mineault, who publish…
Northeastern University computer science student Ryan Amiri used large language models as a "judge" during a co-op at HubSpot to evaluate the quality, relevance and correctness of outputs from popular AI agents, producin…
NVIDIA published a tutorial showing developers how to trace Hermes Agent behavior with NVIDIA NeMo Relay, running two Hermes Agent examples and inspecting the resulting traces for model and tool calls, errors, retries, d…
Merj R&D Director Giacomo Zecchini reported that in 100 purpose-built tests, browser agents failed silently on structural, naming, state and feedback defects that accessibility engineering already addresses — including o…
Google researchers introduced Regularized Recursive Self-Improvement of Agent Harnesses (RRSI), a method that constrains an LLM agent's self-editing of its prompts, control flow, tooling, memory and context management to…
An open-source security script called malfilter analyzes npm package metadata — name similarity, download counts, and creation timestamps — to flag likely malicious packages before installation, targeting typo-squatting …
A year-long investigation by Lighthouse Report and European media outlets found that most data centers across Europe refused to disclose their energy and water consumption, despite the EU Energy Efficiency Directive requ…
Researchers at Chalmers University of Technology in Sweden built a closed-loop AI laboratory that autonomously generated and validated new biological discoveries in brewer's yeast (Saccharomyces cerevisiae), combining la…
The White House launched America.gov on Tuesday, an AI-powered federal portal that searches roughly 29,000 government websites, and President Trump signed an executive order establishing the site, describing it as "a res…
A GPU infrastructure company's five-year engineering retrospective traces its evolution from a naive ECS architecture to Knative on Kubernetes, concluding that cold start was the core problem it fought for years. The com…
A developer detailed an architecture for scaling AI audio transcription backends by decoupling FastAPI's web layer from GPU inference using Celery and Redis. The writeup shows that running synchronous Faster-Whisper tran…
CoreWeave launched ARIA, a coding agent built directly into Weights & Biases that runs a full autoresearch loop — forming hypotheses, writing configs, launching experiments through W&B Launch, evaluating results against …
The Small Business Administration's Office of Inspector General found that the SBA failed to classify its Palantir-built AI fraud detection pilot for pandemic loans as a high-impact use case, exempting it from federal ri…
Researchers submitted TokenCast, a method that learns a composable cost representation for each execution segment of an LLM agent run, forecasting token consumption before and during execution without additional LLM call…
Replikativ released Superficie, a bidirectional renderer that prints Clojure forms as Python-like text with infix arithmetic and blocks and reads that text back into the identical forms, enforcing exact roundtripping by …
Qlty released SlopOne, a free research preview AI model that scores code maintainability by combining traditional static analysis statistics with classifications from TypeSafe AI's new Jev decision model, trained on huma…
NVIDIA published its Open Agent Safety Platform on 28 September 2026, a two-part system pairing the Apache 2.0 OpenShell runtime — built by the Gretel team NVIDIA acquired in 2025 — with NVIDIA Sentry, a watchdog that ru…