QUANTSLANT_
Benchmarks

Thinking Machines Lab models, access and benchmark guide

quantslant@ai:~$ cat news_1

› A small model can earn its place by routing easy, repetitive messages.

quantslant@ai:~$ cat news_2

› An Inkling result on a broad benchmark does not prove it can distinguish a library hold from “hold off.” Give the evaluator real-looking English counterexamples with known labels.

quantslant@ai:~$ cat news_3

› The Inkling 2 Small result concerns a general intelligence evaluation.

quantslant@ai:~$ cat news_4

› A library classifier has a narrower job.

quantslant@ai:~$ open --full
Read more

Thinking Machines Lab models, access and benchmark guide

Open article on quantslant.com