Probabilistic Lipschitz Analysis of Neural NetworksArtifact
Thu 19 Nov 2020 05:40 - 06:00 at SPLASH-III - 5 Chair(s): Xavier Rival, Sukyoung Ryu
We are interested in algorithmically proving the robustness of neural networks. Notions of robustness have been discussed in the literature; we are interested in probabilistic notions of robustness that assume it feasible to construct a statistical model of the process generating the inputs of a neural network. We find this a reasonable assumption given the rapid advances in algorithms for learning generative models of data. A neural network f is then defined to be probabilistically robust if, for a randomly generated pair of inputs, f is likely to demonstrate k-Lipschitzness, i.e., the distance between the outputs computed by $f$ is upper-bounded by the k-th multiple of the distance between the pair of inputs. We name this property, probabilistic Lipschitzness.
We model generative models and neural networks, together, as programs in a simple, first-order, imperative, probabilistic programming language, pcat. Inspired by a large body of existing literature, we define a denotational semantics for this language. Then we develop a sound local Lipschitzness analysis for cat, a non-probabilistic sublanguage of pcat. This analysis can compute an upper bound of the ``Lipschitzness'' of a neural network in a bounded region of the input set. We next present a provably correct algorithm, PROLIP, that analyzes the behavior of a neural network in a user-specified box-shaped input region and computes - (i) lower bounds on the probabilistic mass of such a region with respect to the generative model, (ii) upper bounds on the Lipschitz constant of the neural network in this region, with the help of the local Lipschitzness analysis. Finally, we present a sketch of a proof-search algorithm that uses PROLIP as a primitive for finding proofs of probabilistic Lipschitzness. We implement the PROLIP algorithm and empirically evaluate the computational complexity of PROLIP.
Wed 18 NovDisplayed time zone: Central Time (US & Canada) change
17:00 - 18:20 | |||
17:00 20mResearch paper | Abstract Neural Networks SAS Pre-print Media Attached | ||
17:20 20mTalk | Amalgamating Different JIT Compilations in a Meta-tracing JIT Compiler Framework DLS 2020 Link to publication DOI Pre-print Media Attached | ||
17:40 20mResearch paper | Probabilistic Lipschitz Analysis of Neural NetworksArtifact SAS Ravi Mangal Georgia Institute of Technology, Kartik Sarangmath Georgia Institute of Technology, Aditya Nori , Alessandro Orso Georgia Tech Pre-print Media Attached | ||
18:00 20mTalk | Pricing Python Parallelism: A Dynamic Language Cost Model for Heterogeneous Platforms DLS 2020 Dejice Jacob University of Glasgow, UK, Phil Trinder University of Glasgow, Jeremy Singer Glasgow University Link to publication DOI Pre-print Media Attached |
Thu 19 NovDisplayed time zone: Central Time (US & Canada) change
05:00 - 06:20 | |||
05:00 20mResearch paper | Abstract Neural Networks SAS Pre-print Media Attached | ||
05:20 20mTalk | Amalgamating Different JIT Compilations in a Meta-tracing JIT Compiler Framework DLS 2020 Link to publication DOI Pre-print Media Attached | ||
05:40 20mResearch paper | Probabilistic Lipschitz Analysis of Neural NetworksArtifact SAS Ravi Mangal Georgia Institute of Technology, Kartik Sarangmath Georgia Institute of Technology, Aditya Nori , Alessandro Orso Georgia Tech Pre-print Media Attached | ||
06:00 20mTalk | Pricing Python Parallelism: A Dynamic Language Cost Model for Heterogeneous Platforms DLS 2020 Dejice Jacob University of Glasgow, UK, Phil Trinder University of Glasgow, Jeremy Singer Glasgow University Link to publication DOI Pre-print Media Attached |