swiftywizard@discuss.tchncs.de to Programmer Humor@programming.dev · edit-26 hours agoLavalamp too hotdiscuss.tchncs.deimagemessage-square20fedilinkarrow-up1194arrow-down16
arrow-up1188arrow-down1imageLavalamp too hotdiscuss.tchncs.deswiftywizard@discuss.tchncs.de to Programmer Humor@programming.dev · edit-26 hours agomessage-square20fedilink
minus-squareAlex@lemmy.mllinkfedilinkarrow-up41·5 hours agoIf you have ever read the “thought” process on some of the reasoning models you can catch them going into loops of circular reasoning just slowly burning tokens. I’m not even sure this isn’t by design.
minus-squareswiftywizard@discuss.tchncs.deOPlinkfedilinkarrow-up26·5 hours agoI dunno, let’s waste some water
minus-squareSubArcticTundra@lemmy.mllinkfedilinkarrow-up5·4 hours agoI’m pretty sure training is purely result oriented so anything that works goes
If you have ever read the “thought” process on some of the reasoning models you can catch them going into loops of circular reasoning just slowly burning tokens. I’m not even sure this isn’t by design.
I dunno, let’s waste some water
I’m pretty sure training is purely result oriented so anything that works goes