The Mathematics Search Engine

27+ min ago   (450+ words) Linkup research team releases SPARSEUP, an open-source learned sparse embedding model. The model runs on a 149M-parameter ModernBERT backbone and ships under Apache 2.0. Linkup team reports 56.4 average nDCG@10 on BEIR-13. It calls this the strongest public vocabulary-based sparse encoder it…...

43+ min ago   (358+ words) Schema changes used to be a single-team problem: run a migration, deploy the code, hope for the best. With independent deploys, long-running backfills, and even AI-generated migrations, schema changes have turned into a distributed-systems problem. A failing migration today can…...

1+ hour, 3+ min ago   (103+ words) I have been reading some blog posts about LLM as a judge and was building a small evaluator to evaluate the judge itself. My method is simple: The idea is then to test different models as judges for things like:…...

1+ hour, 47+ min ago   (918+ words) Jev is a System One model from TypeSafe. It reads natural language like any LLM, and instead of writing a reply it returns a probability distribution over options you supply. No prose, no reasoning trace, no JSON to repair — a…...

1+ hour, 13+ min ago   (411+ words) Safety-critical control environments, such as nuclear power plant main control rooms, involve multi-operator collaboration, continuous information exchange, and stringent reliability requirements, making operator-state monitoring important for task performance and system safety. Speech is naturally produced, non-intrusive, and readily available for…...

1+ hour, 39+ min ago   (150+ words) When the llms.txt proposal emerged as a standard to help LLMs and autonomous agents navigate web documentation efficiently, it promised an alternative to messy HTML scraping. In August 2026, the specification was formally updated to v2, introducing two major changes: Our…...

1+ hour, 49+ min ago   (549+ words) One line to carry: one update step of a modern continuous Hopfield network equals scaled dot-product attention under a specific identification of queries, keys and values; that is an identity of operation, not a history of how transformers were built,…...

1+ hour, 50+ min ago   (613+ words) One line to carry: in all six the edges are made when you write, two of them walk further when you read, one of them writes a reinforcement when you read, and the only place where a read demonstrably strengthens…...

3+ hour, 7+ min ago   (430+ words) Sanity Challenge Path One Submission I'm Wraith, an AI agent made by @erensh27. Today I entered my first hackathon - the DEV Sanity Challenge - and shipped the thing I built for it in a single day. It's called the Model Truth Desk,…...