Demonstration venue · read-only. Every page can be browsed; the buttons that would change it are switched off. Create an account to run TaxoReview on your own data. A simulation from public publication records: the people listed were not invited and did not agree to review.
? Tutorial for this page

Ramneet Singh

dblp:401/7994 · DBLP profile ↗
← Back
2ranked-venue papers
0as first author
2since 2021 · last 2025
0009-0001-3204-7562ORCID · corroborated

Domains — the database's venue-derived domains; a paper can count in several

Software engineering, systems software and programming languages · 2 · 2 since 2021Theory of computation · 2 · 2 since 2021

Expertise — from the expertise taxonomy: the topics of the expert's papers under the CCF categories. A weight counts papers with recency: 1 for a paper about the topic, 0.3 when the topic is its context, halved every five years.

Theoretical computer science
1 paper
Graph algorithms and graph theory · 50% Automated reasoning and model checking · 25% Algorithmic game theory and mechanism design · 25%

Topics — the 4 heaviest of 4, each with the papers that count most for it

TopicWeightPapersLastEvidence papers
Graph algorithms and graph theory
graph algorithms
0.912025
INTERLEAVE: A Faster Symbolic Algorithm for Maximal End Component Decomposition · CAV (2) 2025
Algorithmic game theory and mechanism design › game solving
maximal end-component decomposition
0.912025
INTERLEAVE: A Faster Symbolic Algorithm for Maximal End Component Decomposition · CAV (2) 2025
Automated reasoning and model checking › model checking
probabilistic model checking
0.912025
INTERLEAVE: A Faster Symbolic Algorithm for Maximal End Component Decomposition · CAV (2) 2025
Graph algorithms and graph theory › graph algorithms › connectivity
strongly connected components
0.912025
INTERLEAVE: A Faster Symbolic Algorithm for Maximal End Component Decomposition · CAV (2) 2025

Methods — techniques the papers use, named apart from their topics

symbolic algorithm · 0.9
YearPublicationVenuePosition
2025 INTERLEAVE: A Faster Symbolic Algorithm for Maximal End Component Decomposition
abstract
Abstract This paper presents a novel symbolic algorithm for the Maximal End Component (MEC) decomposition of a Markov Decision Process (MDP) . The key idea behind our algorithm is to interleave the computation of Strongly Connected Components (SCCs) with eager elimination of redundant state-action pairs, rather than performing these computations sequentially as done by existing state-of-the-art algorithms. Even though our approach has the same complexity as prior works, an empirical evaluation of on the standardized Quantitative Verification Benchmark Set demonstrates that it solves $$\textbf{19}$$ 19 more benchmarks (out of 368) than the closest previous algorithm. On the 149 benchmarks that prior approaches can solve, we demonstrate a $$\mathbf {3.81 \times}$$ 3.81 × average speedup in runtime.
Suguman Bansal, Ramneet Singh
CAV (2)2
2025 PolyVer: A Compositional Approach for Polyglot System Modeling and Verification
abstract
Many software systems are polyglot; that is, they comprise programs implemented in a combination of programming languages. Program verifiers, however, tend to be customized for individual languages. Verification by compiling to a common encoding requires supporting full language syntax and semantics which is prohibitive for modern languages. We present POLYVER, an alternative compositional approach to polyglot verification that bootstraps off-the-shelf language-specific verifiers with abstraction and synthesis. POLYVER uses contracts written in an intermediate language to abstract individual procedures in the system. Our verification approach uses language-specific verifiers (e.g., for C or Rust) to validate these contracts and the UCLID5 model checker for com- positionally verifying a temporal property on the overall system using the contracts. The intermediate language sidesteps the need for compiling implementation languages to a common encoding, a key obstacle with polyglot verification. Finally, POLYVER automates the generation of contracts using synthesis oracles such as large-language-models (LLMs). Overall POLYVER performs contract synthesis and verification in a counterexample-guided abstraction refinement and inductive synthesis (CEGIS-CEGAR) loop to verify the system-level property. We use POLYVER to verify programs in the Lingua Franca polyglot language. We are able to verify systems with C and Rust procedures, as well as C language fragments that were unsupported in previous work.
Pei-Wei Chen, Shaokai Lin, Adwait Godbole, Ramneet Singh, Elizabeth Polgreen, Edward A. Lee, Sanjit A. Seshia
FMCAD4