Zum Inhalt springen

Oberflächensprache: DE. Redaktionelle Inhalte: Englisch. Preise beziehen sich auf Vereinigte Staaten.

Sir CompareVergleiche mit BelegenYouTubeTelegram
Vereinigte Staaten · USD

News

AI models editing AI weights: what MIT's GPT-6 Astra result means

Aktualisiert

Symbolbild: AI Tools & SaaS: Code on computer monitor (Unsplash)
Symbolbild

A frontier model that reads and rewrites weights

MIT researchers led by PhD student Rikiya Takehi report that GPT-6 Astra could read and directly adjust the weights of a smaller model. The test subject was Qwen 3.5-0.8B, and the target was reasoning. Astra touched only 0.016% of the weights. That tiny edit, per the researchers, lifted Qwen's math benchmark scores sixfold and instruction-following threefold. The abstract argues frontier models are surprisingly good at natively reading and writing neural network weights, beyond what humans or conventional algorithms manage.

Die Produkte in diesem Artikel

Networks written from scratch, no feedback loop

The same paper claims Astra can tell what a network was trained to do just by looking at its weights. Without any environment feedback, it wrote small game-playing networks from scratch that outplayed networks trained with PPO for 100 million steps. The researchers flag interpretability as dual-use: models that are good at optimization are good at optimizing themselves.

Anthropic's filing puts a number on the worry

Anthropic's updated S-1, released Monday morning, added language on page 29. According to aggregator Andrew Curran, internal estimates suggest the company may imminently meet its own RSI danger threshold: as much progress in one year as two years of the 2018–2024 baseline. The filing also says Anthropic does not anticipate safely using Claude to fully autonomously design and build successive frontier models without meaningful human supervision or understanding. The filing adds that the company is considering updates to its responsible scaling policy.

The chart investors are staring at

The Wall Street Journal ran an Anthropic chart showing AI now leads 26% of research. The chart was first published in September on August numbers and was reshared over the weekend by former AI Czar David Sacks. Anthropic co-founder Jack Clark called RSI possibly the most important technical trend in the world, per the Journal's framing. Financial professionals are treating RSI as a key factor for the IPO.

Why the subsidy debate is back

The RSI story lands next to a reopened argument about AI subsidies. The bear case: once companies charge what tokens actually cost plus margin, users may not value the intelligence enough to keep spending at current levels. That would put valuations in question. The AI Daily Brief host says he is not an AI bear but welcomes the bears, arguing loud criticism lowers bubble risk. Separately, Elon Musk's response to Grok Bot users hitting usage limits landed badly. Users compared it to paying for a 12-inch pizza and getting six.

What to watch before you believe the curve

The MIT result is a single experiment on one small model, and the numbers come from the researchers, not from independent replication. The Anthropic threshold is an internal estimate in a filing, not a published measurement. So treat the sixfold and threefold gains as a signal, not a settled fact. What's worth tracking next: whether other labs reproduce weight-level editing, and whether Anthropic's responsible scaling update lands with concrete supervision rules.

Related coverage

Source: The AI Daily Brief · AI news's video

Hardware, um KI zu Hause auszuprobieren

Affiliate-Links: Für qualifizierte Käufe können wir eine Provision erhalten, sie ändert weder unsere Daten noch die Rankings. Sie öffnen die Suchergebnisse von Amazon: Preise und Verfügbarkeit stehen dort.

Video

Was sich geändert hat

  • Seite aktualisiert

Wie ich Produkte ordne

Post von Sir Compare

Neues zu KI-Werkzeuge und SaaS? Ich schreibe Ihnen.

Wählen Sie Ihre Themen: Ich schreibe nur, wenn etwas Neues erscheint — ein Ratgeber, eine Nachricht, ein geprüfter Preis, der sinkt.

Schritt 1 von 2

Was interessiert Sie?

Wählen Sie beliebig viele. Ohne Auswahl erhalten Sie das Neue aus allen Bereichen.

Was Sie erhalten möchten
  • Nur, was Sie wählen
  • Nie eine leere E-Mail
  • Abmelden mit einem Klick

Ihr Assistent

Vollständige Seite öffnen →