-
About “having fun”
I asked Kyle: “how do you do your character or story-telling when you build on a concept or a move in a round?” He answered: “to have fun and play around within the moves!”. I take this to heart because I am not used to having fun, as strange as it sounds. And I think…
-
How to wow the crowds
Spectators love SURPRISES! You want the crowds to react to your freestyle, simply throw surprises to them. There are a few ways to build surprises: Importantly, as a reminder to self, expecting hypes in the process of building, as in building to a climax or simply showing off my techniques, will inevitably lead to disappointments…
-
Poppers in Open-style battles
I was told by Chris Kaku a year or 2 ago that poppers are at a disadvantage in open style battles. I think there is truth to it in that the music in open style battles is generally less consistent in rhythm/beats, tho I won’t say less funky, and popping moves are generally less flashy.…
-
Study Probabilistic Modeling
I have studied everything until 2.1. Methods such as maximum likelihood and Bayesian parameter estimation are used to learn probabilistic models. Probabilistic modeling deals with uncertainties and is used to predict probability of future events. To paraphrase, probabilistic models output a probabilistic distribution over possible answers.
-
Amortized inference
“This raises the problem of amortized inference: howto flexibly reuse inferences so as to answer a variety of related queries” What does this mean exactly? And why does the reward function representing the prior x likelihood makes it so? “This makes GFlowNets amortized probabilistic inference machines that can be used both to sample latent variables…
-
Stat/prob/ml concepts
Latent and observable variables In statistics, latent variables (from Latin: present participle of lateo, “lie hidden”) are variables that can only be inferred indirectly through a mathematical model from other observable variables that can be directly observed or measured.[1] Latent variables may correspond to aspects of physical reality. These could in principle be measured, but…
-
GFlowNets study notes
GFlowNet is really MANY THINGS IN ONE THING. And I need to put all aspects together in order to fully understand it. Like, how is this a stochastic policy in the light of it being a generative model i.e., P(y,x)? Generative active learning, reinforcement learning, stochastic policy, generative model, energy-based probabilistic modelling, variational models and…
-
How do GFlowNets relate to attention or human thinking?
How is that related to attention? It makes sense to consider deterministic transitions when the policy is an internal policy, not operating in some stochastic external environment, but instead performing computational choices (like what to attend to, what computation to perform, what memory to retrieve, etc). Immediately, GFlowNets are motivated by the kind of internal…
-
Subscribe
Subscribed
Already have a WordPress.com account? Log in now.
Gang Fang's Blog 