AI Alignment Forum
(Originally posted to the Deepmind Blog) Specification gaming is a behaviour that satisfies the literal specification of an objective without achieving the intended outcome. We have all had experiences with specification gaming, even if not by this name. Readers may have heard the myth of King Midas and the golden touch, in which the king asks that anything he touches be turned to gold - but soon
Stochastic Gradient Descent as Approximate Bayesian Inference
Stochastic Gradient Descent with a constant learning rate (constant SGD) simulates a Markov chain with a stationary distribution. With this perspective, we derive several new results. (1) We show that constant SGD can be used as an approximate Bayesian posterior inference algorithm. Specifically, we show how to adjust the tuning parameters of constant SGD to best match the stationary distribution
Big Data時代の大規模ベイズ学習�-Stochastic Gradient Langevin Dynamics�を中心として
Big Data時代の大規模ベイズ学習�-Stochastic Gradient Langevin Dynamics�を中心として
Bridging the Gap between Stochastic Gradient MCMC and Stochastic Optimization
Stochastic gradient Markov chain Monte Carlo (SG-MCMC) methods are Bayesian analogs to popular stochastic optimization methods; however, this connection is not well studied. We explore this relationship by applying simulated annealing to an SGMCMC algorithm. Furthermore, we extend recent SG-MCMC methods with two key components: i) adaptive preconditioners (as in ADAgrad or RMSprop), and ii) adapti
Langevin Monte Carlo法には棄却が必要か
機械学習系のベイズっぽい論文を読んでいると、SGLD [WT11] やSGRLD [PT13] といった文字列を見ることがあります。これらが何かというとマルコフ連鎖モンテカルロ法 (MCMC) の一種で、正規化定数がわからない高次元の確率分布からのサンプリングを得たい場合などに使われます。 アルゴリズムの位置づけとしては、 Langevin Monte Carlo (LMC) とか Langevin Dynamics などという名前で呼ばれている既存アルゴリズムがまずあり、それに伴う勾配計算をサブサンプリングを利用して簡略化したもの という感じです。勾配降下法 (GD) を確率的勾配降下法 (SGD) に拡張することにインスパイアされているのだったと思います。LMCのモチベーションとしてよく言われるのは、LMCでは (Metropolis–Hastings法に見られるような) 棄却ステップ
Preconditioned Stochastic Gradient Langevin Dynamics for Deep Neural Networks
Effective training of deep neural networks suffers from two main issues. The first is that the parameter spaces of these models exhibit pathological curvature. Recent methods address this problem by using adaptive preconditioning for Stochastic Gradient Descent (SGD). These methods improve convergence by adapting to the local geometry of parameter space. A second issue is overfitting, which is typ
1
リリース、障害情報などのサービスのお知らせ
最新の人気エントリーの配信
j次のブックマーク
k前のブックマーク
lあとで読む
eコメント一覧を開く
oページを開く
Stochastic Gradient and Langevin Processes
[1509.01631] Stochastic gradient variational Bayes for gamma approximating distributions
Everything that Works Works Because it's Bayesian: Why Deep Nets Generalize?
Energy-Based Generative Adversarial Network
https://www.jstage.jst.go.jp/article/tjsai/31/6/31_AI30-C/_pdf
http://www.jmlr.org/papers/volume17/teh16a/source/teh15a.pdf
Stochastic Gradient Riemannian Langevin Dynamics on the Probability Simplex