TokenPost topic

OLMo-3.1-32B-Think

1 TokenPost story is tagged OLMo-3.1-32B-Think. The newest appear first.

Latest OLMo-3.1-32B-Think stories

• Blocking Common Opening Tokens and Adding an Example Raised Reasoning Compliance in Two AI Models.

  Blocking “Okay” and “Alright” while providing an example raised average written-reasoning compliance by 41.2 percentage points for Qwen3-32B and 34.5 percentage points for OLMo-3.1-32B-Think; the scores did not measur…

  Nov 3, 2026, 10:19 PM EST · Technology