Di Wang 王迪

Doctoral Student in CS

Carnegie Mellon University

Biography

I am a second-year doctoral student in computer science at Carnegie Mellon University. I am advised by Prof. Jan Hoffmann. I am broadly interested in programming languages and software engineering, especially probabilistic programming, quantitative program analysis, higher type theory, and program synthesis.

I completed my undergraduate at Peking University, China where I worked with Prof. Yingfei Xiong on summarization techniques to analyze programs sharing big libraries.

Here is my Curriculum Vitae.

News

I’m excited that our paper Type-Guided Worst-Case Input Generation with Jan Hoffmann has benn accepted to POPL 2019.
Check out my talent show Chengdu on SCS Day 2018!
I’m excited that our paper PMAF: An Algebraic Framework for Static Analysis of Probabilistic Programs (with Jan Hoffmann and Tom Reps) has been conditionally accepted to PLDI 2018.

Recent Publications

Di Wang, Jan Hoffmann. Type-Guided Worst-Case Input Generation. In POPL, 2019.

Di Wang, Jan Hoffmann, Thomas Reps. PMAF: An Algebraic Framework for Static Analysis of Probabilistic Programs. In PLDI, 2018.

Preprint PDF Slides

Di Wang, Jan Hoffmann, Thomas Reps. A Denotational Semantics for Nondeterminism in Probabilistic Programs. Working paper, 2018.

Peng Wang, Di Wang, Adam Chlipala. TiML: A Functional Language for Practical Complexity Analysis with Invariants. In OOPSLA, 2017.

Preprint Code Web Interface GitHub

Hao Tang, Di Wang, Yingfei Xiong, Lingming Zhang, Xiaoyin Wang, Lu Zhang. Conditional Dyck-CFL Reachability Analysis for Complete and Efficient Library Summarization. In ESOP, 2017.

Preprint Artifact

Recent & Upcoming Talks

PMAF: An Algebraic Framework for Static Analysis of Probabilistic Programs

Apr 20, 2018 12:00 PM

PLunch

Slides

TiML: A Functional Language for Practical Complexity Analysis with Invariants

Feb 9, 2018 12:00 PM

PLunch

Slides

Recent Posts

Nov 3, 2018 3 min read Problem Solving

Using FFT to Speed Up DP

Problem link: Counting Road Networks | HackerRank. You are supposed to count the number of connected undirected labeled graphs with $n$ vertices. Algorithms with $O(n \log^2 n)$ time complexity are preferable. A Dynamic-Programming Algorithm Let $f(n)$ be the answer for $n$. The first idea to compute $f(n)$ is subtracting the number of disconnected graphs from the total number. The total number of size-$n$ graphs is $g(n) := 2^{\binom{n}{2}}$. How to count the disconnected graphs?

CONTINUE READING

Feb 8, 2018 3 min read Programming

Nondeterministic Interpretation

Suppose you have a toy specification with built-in nondeterminism, and you want to generate answers with respect to the specification: type exp = EInt of int | EPair of exp * exp | ENdet of exp * exp type ans = VInt of int | VPair of ans * ans For example, from the following specification EPair (ENdet (EInt 5) (EInt 6)) (ENdet (EInt 7) (EInt 8)) you might want to generate a bunch of possible answers:

CONTINUE READING