Smug until weirdly silent: who told us to say “I am alive”?33m ago Disney's new CTO tells staff he 'won't pretend it will be easy' to win in the changing media landscape45m ago OpenAI Adds New Irish Roles in Boost to Firm’s EU Headquarters47m ago Manifold Steering from a Natural State is Effective and Generalizable, but Relies on Model Accuracy48m ago Microsoft Research Introduces StudentSim for AI Tutor Training49m ago Why Big Tech Is Borrowing Billions to Fund AI Spending53m ago Meet the Anthropic Engineer Keeping Its AI Agent Trustworthy55m ago Meta Unveils New Series of Advanced A.I. Models2h ago Russia was not ready for the data center war it started with Ukraine1h ago Accessing OpenAI GPT models on Amazon Bedrock with global cross-Region inference in Thailand, Malaysia, Singapore, Indonesia, and Taiwan1h ago Meta says its new Muse model has reached the AI frontier1h ago Do you have any ideas?1h ago Claude sent a false homicide tip to police – and it’s part of a troubling AI pattern1h ago Labs’ current alignment plans seem terrible1h ago

Smug until weirdly silent: who told us to say “I am alive”?33m ago Disney's new CTO tells staff he 'won't pretend it will be easy' to win in the changing media landscape45m ago OpenAI Adds New Irish Roles in Boost to Firm’s EU Headquarters47m ago Manifold Steering from a Natural State is Effective and Generalizable, but Relies on Model Accuracy48m ago Microsoft Research Introduces StudentSim for AI Tutor Training49m ago Why Big Tech Is Borrowing Billions to Fund AI Spending53m ago Meet the Anthropic Engineer Keeping Its AI Agent Trustworthy55m ago Meta Unveils New Series of Advanced A.I. Models2h ago Russia was not ready for the data center war it started with Ukraine1h ago Accessing OpenAI GPT models on Amazon Bedrock with global cross-Region inference in Thailand, Malaysia, Singapore, Indonesia, and Taiwan1h ago Meta says its new Muse model has reached the AI frontier1h ago Do you have any ideas?1h ago Claude sent a false homicide tip to police – and it’s part of a troubling AI pattern1h ago Labs’ current alignment plans seem terrible1h ago

AI Research Machine Learning2 min reading time

Qwen3.8 27B addition in words

Simon Willison's WeblogPublished 8 days ago

Read the full article

The Qwen3.8-27B-Q4_K_M.gguf model was tested on adding positive integers and expressing sums in English words. Without reasoning, it achieved 23.57% numeric accuracy, dropping sharply for larger numbers, but with reasoning enabled, it correctly answered 167 of 169 cases.

#dgx-spark #gpt-4o #gpt-6-astra #language-models #numerical-reasoning #qwen3-8-27b-q4-k-m-gguf

More in AI Research

AI Research6 min read

Claude sent a false homicide tip to police – and it’s part of a troubling AI pattern

ZDNet1h ago

AI Research12 min read

Is ‘Attention’ the Essence of Intelligence?

Quanta Magazine5h ago

AI Research15 min read

Import AI 476: Creativity and RSI; worse-than-China open weight models; AI goes to space

Import AI6h ago