Qwen3.8 27B addition in words
Research: Qwen3.8 27B addition in words
Proof1 independent outlet
Key points
- Colin Frasier posted on Bluesky about an experiment he ran over two years ago using GPT-4o to see how well it could "compute the sum but return the answer in words" across increasingly large numbers.
- I'm confident GPT-4o didn't cheat and use a calculator, especially since it got so many of the calculations wrong, but I was inspired to run the experiment again on local hardware (a DGX Spark) to explore the effect in a fully controlled environment.
- I pasted his image into a Codex Remote session (GPT-6 Astra) and had it run the same experiment using Qwen3.8-27B-Q4KM.gguf.
- Here's a version of the report that includes the reasoning traces from some of those larger calculations, which include text like this:
Sources (1)
- [1]Qwen3.8 27B addition in wordsSimon Willison's Weblog · Oct 4, 11:34 PM
Research: Qwen3.8 27B addition in words
Colin Frasier posted on Bluesky about an experiment he ran over two years ago using GPT-4o to see how well it could "compute the sum but return the answer in words" across increasingly large numbers.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 4, 2026nerkyor/Qwen3.8-27B-Coder390-EfficientThink-Opus5.5-GPT6Astra-Grok4.7-DSV4Pro-K3-SFT-RLOO-MTP-DFlash2
- Oct 3, 2026[AINews] not much happened today
- Oct 1, 2026Claude-shaped science
- Sep 30, 2026[AINews] OpenAI DevDay 2026: Dots, 6.1 Sol, Ultrafast, Decisions API, Agents API, Spaces, Marketplace, and 1.2 Billion ChatGPT WAU
- Sep 29, 2026DevDay 2026 Recap
- Sep 28, 2026Holo4: powering generalist computer-use agents
