NotEvenTrying⏧
@rizkidotme
Followers
403
Following
27K
Statuses
7K
realtime learning, real life performer. please, remind me if i'm not paying attention to things I care about! SLAM-ing the edge of chaos
placeholder
Joined March 2015
RT @ArtemKRSV: Brain circuits, gene networks, social media - they all share the same underlying pattern of connections called small-world g…
0
25
0
RT @dileeplearning: Exciting! 'Space is a latent sequence' & CSCG theory of hippocampus we developed at @GoogleDeepmind and @vicariousai ge…
0
17
0
RT @Plinz: What Tesla foresaw was not just the cellphone, it was AI synchronizing humanity and moving into the ambient space
0
12
0
RT @dwarkesh_sp: The @JeffDean & @NoamShazeer episode. We talk about 25 years at Google, from PageRank to MapReduce to the Transformer to…
0
192
0
RT @BartBussmann: Do SAEs find the ‘true’ features in LLMs? In our ICLR paper w/ @neelnanda5 we argue no The issue: we must choose the num…
0
36
0
RT @HazanPrinceton: V excited about a breakthrough in learning linear dynamical systems and sequence prediction, w. my brilliant postdoc,…
0
13
0
RT @repligate: Bullshit. The reason is not boring or complicated or technical (requiring domain knowledge) Normies are able to understand…
0
33
0
these models just like me fr
Through failure mode analysis, we identify a new form of memorization -- models may memorize the problem-solving skills from the training set and blindly apply them without judging whether the modified settings are suitable or not. 🧵 [4/n]
0
0
0
RT @eshear: A lot of my best ideas are come from when someone else says something really interesting but kind of off-topic and it turns out…
0
19
0
@wordgrammer @IvanVendrov do you think they'll update their conceptions about ethics for human?
0
0
0
RT @IvanVendrov: why don't philosophers talk about computational constraints more? utilitarianism make sense - if you have infinite compute…
0
35
0
@adamnemecek1 @__paleologo @marc_tsundoku overtime "executing" partition function to a system will become less complex than previous one.
0
0
1
@eudaemonea @packyM here's mine: the reward function of a system shall not unbounded and have positive feedback loop dynamics, otherwise it'll just "shoot itself in the foot." Any stable and self-sustainable system we find in nature (including ourself) only have "hedonic treadmill" when too stable.
0
0
1
@adamnemecek1 @__paleologo @marc_tsundoku all of these phenomena are complementary of second law of thermodynamics actually
1
0
2
