https://nosible.com/blog 2026-06-18 https://nosible.com/blog/the-contrastive-geometry-of-risk 2026-06-18 https://nosible.com/images/2026/06/walkthrough_step9_number_line.png Two Tricks for Turning Sentence Embeddings into Clean Features A ten-step, training-free walkthrough that turns a frozen OpenAI text embedding into clean classifications: a multiclass relevance score sorts events into local, national, and global buckets, and a contrastive binary score splits systemic from idiosyncratic risk. Verified on real warnings from NOSIBLE World, the geometry matches Google's gemini-2.5-flash while staying deterministic, auditable, and effectively free. https://nosible.com/blog/an-embedding-based-approach-to-trade-and-economic-policy-uncertainty 2026-06-17 https://nosible.com/images/2026/06/nosible-tpu-daily.png An Embedding-Based Approach to Trade and Economic Policy Uncertainty The Fed's Trade Policy Uncertainty index counts keywords across seven newspapers. We rebuilt it from 14.9 million NOSIBLE World events using only embeddings and five sentences, no keywords. It matches the published benchmark at 0.87 on monthly levels and 0.82 on monthly changes, as closely as the two official versions match each other. The same method, extended to sixty sentences, rebuilds the broader Economic Policy Uncertainty index and its national-security and healthcare categories. https://nosible.com/blog/turning-news-into-a-risk-on-risk-off-equity-signal 2026-06-16 https://nosible.com/images/2026/06/news-stress-overlay-hero.png Turning News into a Risk-On/Risk-Off Equity Signal We built a risk-on/risk-off trading signal from the NOSIBLE event database that measures how much of the global news flow is about market-stress themes, holding equities when that reading is low and moving to T-bills when it spikes. Selected on 2010 to 2013 and tested on an untouched 2015 to 2026 window, it held the S&P 500's buy-and-hold return (+254% versus +269%) while cutting the maximum drawdown from −34% to −18% and raising the Sharpe ratio from 0.64 to 0.89. The same rule transfers unchanged to the Nasdaq and the Russell 2000. https://nosible.com/blog/rebuilding-the-geopolitical-risk-index-from-nosible-world 2026-06-06 https://nosible.com/images/2026/06/nosible-gpr-vs-published.png We Rebuilt the Geopolitical Risk Index with Nosible World Markets move on geopolitics, but risk models cannot read the news. We turned 13.2 million news events into a geopolitical risk signal, matched the Federal Reserve benchmark, and broke it down by country, by country pair, and into an oil supply-risk signal. https://nosible.com/blog/fast-enough-to-matter-productionizing-tiny-transformers-for-signal-extraction 2025-12-12 https://nosible.com/blog/illustrations/the-sprinter.png Matching GPT-5.1 at Financial Sentiment with Active Learning and Qwen3 Here's how we fine-tuned Qwen3 0.6B to beat FinBERT and match GPT-5.1 accuracy. Complete with open-source models, datasets, and training scripts. Spoiler alert: active learning is all you need. https://nosible.com/blog/can-faceted-search-at-web-scale-self-organize 2025-10-16 https://nosible.com/blog/illustrations/vortex.png Can Faceted Search at Web-Scale Self Organize? Can Faceted Search at Web-Scale Self Organize? As it turns out, yes it can! In this post we outline our new and improved adaptive named entity tagging system! https://nosible.com/blog/introducing-cybernaut-1-agentic-search-with-mcts 2025-08-26 https://nosible.com/blog/illustrations/cyber.png Introducing Cybernaut-1: Agentic Search using MCTS Cybernaut-1 combines our powerful hybrid-3 search algorithm with LLM-guided Monte Carlo Tree Search to deliver world class search results on difficult queries. https://nosible.com/blog/the-road-to-cybernaut-1 2025-08-20 https://nosible.com/blog/illustrations/track.png The Road to Cybernaut-1: Rebuilding Search for AI AI needs its own search engine. This is how we’re rebuilding search for AI -- and the road to Cybernaut-1, the first high-trust agentic search engine. https://nosible.com/blog/ensemble-and-distil 2024-02-06 https://nosible.com/blog/illustrations/cube.png A Pattern for Scaling the Value Proposition of LLMs: Ensemble and Distil 🚀 We introduce the ensemble and distil data pattern and use it to fit an ordinary least squares linear regression that outperforms GPT-4 at financial news sentiment classification using sentence transformer embeddings as features. https://nosible.com/blog/news-sentiment-showdown-who-checks-vibes-best 2024-01-28 https://nosible.com/blog/illustrations/eye.png News Sentiment Showdown: Who Checks Vibes Best? A comparison of sentiment classifications made by TextBlob, VADER, Flair, SigmaFSA, FinBERT, FinBERT-Tone, Text-Bison, Text-Unicorn, Gemini-Pro, GPT-3.5, GPT-4, and GPT-4-Turbo. We look at accuracy, time, and cost and include a dataset of 10,368 labelled news stories (with code) for our followers. https://nosible.com/blog/using-vector-search-to-see-signals-in-company-news 2024-01-21 https://nosible.com/blog/illustrations/inspect.png Using Vector Search to See Signals in Company News How we use vector search to extract investment signals from a multi-terabyte company news dataset that currently contains over 55 million embeddings, 150+ million sentences, 4+ billion words, and 5+ billion GPT tokens.