Ages2Come
Menu
View all results →
Categories
⚡ Breaking
Incumbent Sen. Ed Markey wins Massachusetts' Democratic primary for US Senate Apple Maps changes Lake Ontario to Lake America after Trump order Larry Page’s flying car company Pivotal loses its CEO Bill Simmons’s Co-Hosts Are Tired of His A.I. Antics SEC proposes transfer agent rule, sets event to figure out round-the-clock U.S. trading China dissented from G20 statement opposing 'cheap exports' flooding market, Bessent says Tim Cook to be paid $47mn as Apple’s executive chair Are all TVs at risk from burn-in? Here's how to avoid the problem altogether At G20 Meeting, Scott Bessent Accuses China of Flooding the World With Cheap Exports Will self-flying planes transform the skies? Incumbent Sen. Ed Markey wins Massachusetts' Democratic primary for US Senate Apple Maps changes Lake Ontario to Lake America after Trump order Larry Page’s flying car company Pivotal loses its CEO Bill Simmons’s Co-Hosts Are Tired of His A.I. Antics SEC proposes transfer agent rule, sets event to figure out round-the-clock U.S. trading China dissented from G20 statement opposing 'cheap exports' flooding market, Bessent says Tim Cook to be paid $47mn as Apple’s executive chair Are all TVs at risk from burn-in? Here's how to avoid the problem altogether At G20 Meeting, Scott Bessent Accuses China of Flooding the World With Cheap Exports Will self-flying planes transform the skies?
AI / Automotive Automotive

Researchers say they trained a foundation model from scratch for about $1,500

Source: ventureBeat June 10, 2026 · 14:52
Training a foundation LLM from scratch costs millions and requires internet-scale data — which is why most enterprises don't bother. Sapient thinks it has a cheaper path.To overcome this brute-force scaling dogma, researchers at Sapient developed HRM-Text, which replaces standard Transformers with a highly sample-efficient Hierarchical Recurrent Model (HRM), an architecture they first introduced last year.HRM decouples computation into slow-evolving strategic and fast-evolving execution layers. Instead of brute-force autoregressive prediction on raw text, HRM-Text trains exclusively on instruction-response pairs. This is close to real-world enterprise settings, where users usually expect a targeted answer to a specific task.The researchers were able to train a 1B-parameter HRM-Text from sc…
Read original article ← Back to Automotive

The original article opens on the publisher's website.

More from Automotive

View topic →