Open Doctrines
open-doctrines.bsky.social
An official account of Open Doctrines Open Doctrines website: https://opendoctrines.pages.dev/ Open Doctrines itch.io: https://pr1nted.itch.io/open-doctrines
👤 Joined on 15 September 2026
2
Followers
1
Following
7
Total Posts
📜 Kullanıcı Geçmişi
- 15 Eyl 2026, 20:07
🎉 Bluesky'e katıldı - İlk kullanıcı adı:
@open-doctrines.bsky.social
🚫 Blokladığı Kullanıcılar
❤️ Beğendiği Gönderiler
1.3.1b — running on a phone, fog of war introduction, and a new map - Open Doctrines pr1nted.itch.io/open-doctrin...
Show PostOpen Doctrines is in beta - 1.3.0b - Open Doctrines pr1nted.itch.io/open-doctrin...
Show Post
[1/2] Thanks. A related result: I used the FAFB graph as a frozen reservoir (readout only) on a binary delayed-cue task. Real wiring 94.8%, degree-preserving rewiring 57.6%, same-density random 52.8%. All three graphs had uniform row-normalized weights and the same gain, so only topology differed.
Show Post[2/2] One thing I like about permuting the postsynaptic column: each presynaptic neuron keeps its sign and its outgoing synapse counts, so Dale's law survives. My control dropped signs and weights on both sides, so yours is the stricter test. I look forward to that comparison.
Show PostI read the preregistration. Good to see the shuffled-connectome control (postsynaptic column permuted) fixed before the runs. Has it been run yet? I'd like to know whether the real wiring beats the shuffled one, separately from whether it beats random-legal play.
Show Post