Patrick Hammer

OpenNARS v1.5.5 had about 3 more multistep example which would be nice to have as multistep testcase for v3.0.2.

At the workshop Hugo raised an important issue about how we decide what elements to keep in the input buffer. Since the structure was not simple, our answers were a...

Don't treat intervals in questions as placeholders, "interval-project" answers to the question, answering with the most similar high-confident answers.

Instead of taking the order of magnitude as the default tolerance (which will still be used as initial value) to generate negative feedback on disappointment of an anticipation, the tolerance...

Documentation (nars core)

6

The job is here: 1. Remove all dead code from nars.core. 2. Remove unnecessary new lines, the policy should be as it was in 1.5.5 here. 3. Add comprehensive documentation...

This is the list of tests that do not pass: nars_seqABC (previous anticipation handling was maybe better) nal9.0 (NAL9 needs some work) nal9.1 (NAL9 needs some work) nal6.23 (second level...

Image scratchpad

12