Tootfinder

#Gaming #CyberPunk2077

Konsequent _drauf geguckt_ kommt raus, dass das Geld was die Anthropic-Token gekostet hätten, um die Bugs zu finden, wenn man es denn verwenden würde um menschliche Sicherheitsforscher zu bezahlen, grob den gleichen Effekt hätte*.
Aber fächelt euch ruhig weiter Luft zu…
https://social.bund.de/@bsi/1163802904

RE: https://social.coop/@cwebber/116110194513314869
seeing a future of personal computing with people running open source operating systems on their outdated cyberdecks with scrounged-up hardware from the detritus that datacenters leave behind when the next generation of GPUs and storage becomes available (but only to large corporations)
very cyberpunk

Had fun playing #Cyberpunk2077 after a LOOOOOOOONG time.
my netrunner V has either 2 modes - quickhacks or guns. No middle ground. (Is there anything else I wonder?)
I need to re-learn how to grab people from behind so I can put them to sleep. I forgot the keybinding for it.
Anyways, it was 3 hours well spent.

Symbolbild Symbolpolitik.
(Die Regelung ist gut. Macht Österreich schon seit vielen Jahren so. Wollte ich, seit ich davon wusste, schon immer hier haben. An den gestiegenen Preisen der letzten Tage in Deutschland wird das trotzdem nichts ändern.)
Edit: "Pending quote approval" ist ja auch geil. Link, just in case. https://

Rethink Efficiency Side of Neural Combinatorial Solver: An Offline and Self-Play Paradigm
Zhenxing Xu, Zeyuan Ma, Weidong Bao, Hui Yan, Yan Zheng, Ji Wang
https://arxiv.org/abs/2602.20730 https://arxiv.org/pdf/2602.20730 https://arxiv.org/html/2602.20730
arXiv:2602.20730v1 Announce Type: new
Abstract: We propose ECO, a versatile learning paradigm that enables efficient offline self-play for Neural Combinatorial Optimization (NCO). ECO addresses key limitations in the field through: 1) Paradigm Shift: Moving beyond inefficient online paradigms, we introduce a two-phase offline paradigm consisting of supervised warm-up and iterative Direct Preference Optimization (DPO); 2) Architecture Shift: We deliberately design a Mamba-based architecture to further enhance the efficiency in the offline paradigm; and 3) Progressive Bootstrapping: To stabilize training, we employ a heuristic-based bootstrapping mechanism that ensures continuous policy improvement during training. Comparison results on TSP and CVRP highlight that ECO performs competitively with up-to-date baselines, with significant advantage on the efficiency side in terms of memory utilization and training throughput. We provide further in-depth analysis on the efficiency, throughput and memory usage of ECO. Ablation studies show rationale behind our designs.
toXiv_bot_toot