Science is best shared!
- Source
- thinkingmachines.ai
- Author
- Thinking Machines
- Date
Context
Thinking Machines's November 2025 call for Tinker community projects gives contributors a concrete standard for work it wants to feature: an explanatory write-up, preferably accompanied by open-source code, with clear and comparisons. The call welcomes replications and infrastructure contributions alongside original research, custom models, datasets and libraries.
That makes a carefully documented reproduction a relevant submission even without a novel algorithm, since the call favors transparent results over novelty. Contributors are asked to explain assumptions and changes, include raw outputs alongside summaries, and compare new methods with simpler baselines so others can assess what the experiment establishes.
- eval — A repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.
- context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
- distillation — Training a small, cheap model to imitate a big one's outputs, keeping much of the capability at a fraction of the cost.
Checking sign-in…
Loading comments…