arXiv subjects

Chih-Duo Hong

Publications and source records attributed to Chih-Duo Hong.

15 recordsLinked to original sources

arXiv2026

ConcoLixir: Reactive LLM Discovery Oracles for Python Concolic Testing

Concolic testing combines concrete execution with symbolic constraint solving, but Python programs expose recurring limits. Library calls can cause symbolic variables to downgrade to concrete values. Regular expressions, checksums, parsers, and other semantic operations can be hard to solve, and exploration can plateau on already covered paths. We present ConcoLixir, a reactive LLM extension for Python concolic execution. The LLM acts as a discovery oracle, not a replacement for the solver or a correctness oracle. It generates initial seeds, proposes concrete inputs after solver failures, and targets uncovered code when coverage stalls. Each candidate is executed concolically, and only observed coverage and collected path constraints guide later exploration. Across synthetic, real-world, and library targets, ConcoLixir improves mean line coverage over the baseline concolic tester without an LLM oracle by 8.6, 15.1, and 17.0 percentage points. The gains are strongest near semantic barriers and library boundaries, and the full evaluation costs \$1.63 in API charges. These results show that bounded LLM discovery can complement symbolic reasoning without replacing it.

cs.SE↗

arXiv2026

Signature filtering: a lightweight enhancement for statistical watermark detection in large language models

Statistical watermarks help organizations attribute large language model (LLM) outputs, yet existing detectors often struggle when watermark signals are weak, texts are repetitive, or watermarks are edited. We propose signature filtering, a detection-time module that enhances watermark detection without modifying watermark embedding and text generation. It learns a small set of ``signature'' tokens whose presence makes watermark tests unreliable, and removes these tokens before detection. The signatures are obtained by solving a mixed-integer linear program on a small training set, with constraints that maximize the true positive rate. We additionally derive finite-sample and asymptotic bounds under several attacker models (color-blind, color-adaptive, and distributionally correlated). On four well-known watermark families (Kgw, Sweet, Unigram, Exp), four benchmark corpora (C4, MBPP, HumanEval, Code-Search-Net), and six LLMs (Opt-1.3b, Opt-6.7b, Llama2-13b, Llama3.1-8b, Qwen2.5-14b, Phi-3-medium-14b), 2- and 3-gram signatures raise detection rates in weak-signal and low-entropy settings from 8~31% without filtering to 78~99% with filtering, while keeping false positives controllable and often negligible. In stress tests where we scramble sentences and perturb 25~50% of tokens by dilution, deletions, and substitutions, 2-gram filters for Kgw-style watermarks preserve most of the clean-text detection gains, often matching or outperforming the advanced WinMax watermark detector. Signature filtering thus provides a simple, scalable, and model-agnostic add-on to strengthen watermark-based provenance checks for LLM text in information processing workflows.

cs.LG↗

arXiv2026

Robustness Verification of Recurrent Neural Networks with Abstraction Refinement

Certified local robustness verification for recurrent neural networks (RNNs) is challenging because approximation errors introduced by nonlinear relaxations can propagate through recurrent connections and accumulate over time. As a result, scalable linear bound propagation methods often become overly conservative and fail to certify inputs that are in fact robust, especially when many pre-activation intervals cross zero. We propose an abstraction-refinement framework for RNN verification that partitions such intervals to remove the dominant relaxation error: on each refined branch, ReLU becomes exact, and smooth activations such as tanh and sigmoid admit substantially tighter linear envelopes. To control the combinatorial cost of splitting in long sequences, we introduce a SHAP-guided timestep selection strategy that ranks hidden states by their contribution to the verification objective and refines only the most critical timesteps in temporal order. Experiments on CIFAR10 and MNIST stroke benchmarks demonstrate consistent improvements in verification success and robustness-margin tightness over abstraction-only baselines, while exposing clear runtime trade-offs between ReLU and tanh models.

cs.LG↗

arXiv2025

Extracting Robust Register Automata from Neural Networks over Data Sequences

Automata extraction is a method for synthesising interpretable surrogates for black-box neural models that can be analysed symbolically. Existing techniques assume a finite input alphabet, and thus are not directly applicable to data sequences drawn from continuous domains. We address this challenge with deterministic register automata (DRAs), which extend finite automata with registers that store and compare numeric values. Our main contribution is a framework for robust DRA extraction from black-box models: we develop a polynomial-time robustness checker for DRAs with a fixed number of registers, and combine it with passive and active automata learning algorithms. This combination yields surrogate DRAs with statistical robustness and equivalence guarantees. As a key application, we use the extracted automata to assess the robustness of neural networks: for a given sequence and distance metric, the DRA either certifies local robustness or produces a concrete counterexample. Experiments on recurrent neural networks and transformer architectures show that our framework reliably learns accurate automata and enables principled robustness evaluation. Overall, our results demonstrate that robust DRA extraction effectively bridges neural network interpretability and formal reasoning without requiring white-box access to the underlying network.

cs.AI↗

arXiv2025

Influence-Guided Concolic Testing of Transformer Robustness

Concolic testing for neural networks alternates concrete execution with constraint solving to search for inputs that flip model decisions. We present a concolic tester for Transformer classifiers that uses SHAP estimates to rank pending path predicates by their impact on the current prediction. To support self-attention with multiple heads in execution backed by SMT solving, we implement attention semantics in pure Python that are compatible with the solver and make the softmax boundary explicit by concretizing exponentiation arguments. We evaluate our method on CIFAR-10 across three compact Transformer classifiers, ResNet18, and VGG16 under a one-pixel budget and a 900s horizon. Across the 500 model--input pairs in this matched comparison, our method achieves 60% success, compared with 15% for a differential evolution baseline that treats the model as a black box. In the primary two-layer Transformer branch-ordering study, SHAP-based predicate prioritization raises success from 56% to 60% and reduces median attack time by 51%. These results show that influence-guided path exploration can make concolic testing a practical way to find adversarial examples in Transformer models.

cs.SE↗

arXiv2025

Concolic Testing on Individual Fairness of Neural Network Models

This paper introduces PyFair, a formal framework for evaluating and verifying individual fairness of Deep Neural Networks (DNNs). By adapting the concolic testing tool PyCT, we generate fairness-specific path constraints to systematically explore DNN behaviors. Our key innovation is a dual network architecture that enables comprehensive fairness assessments and provides completeness guarantees for certain network types. We evaluate PyFair on 25 benchmark models, including those enhanced by existing bias mitigation techniques. Results demonstrate PyFair's efficacy in detecting discriminatory instances and verifying fairness, while also revealing scalability challenges for complex models. This work advances algorithmic fairness in critical domains by offering a rigorous, systematic method for fairness testing and verification of pre-trained DNNs.

cs.LG↗

arXiv2025

Probabilistic Bisimulation for Parameterized Anonymity and Uniformity Verification

Bisimulation is crucial for verifying process equivalence in probabilistic systems. This paper presents a novel logical framework for analyzing bisimulation in probabilistic parameterized systems, namely, infinite families of finite-state probabilistic systems. Our framework is built upon the first-order theory of regular structures, which provides a decidable logic for reasoning about these systems. We show that essential properties like anonymity and uniformity can be encoded and verified within this framework in a manner aligning with the principles of deductive software verification, where systems, properties, and proofs are expressed in a unified decidable logic. By integrating language inference techniques, we achieve full automation in synthesizing candidate bisimulation proofs for anonymity and uniformity. We demonstrate the efficacy of our approach by addressing several challenging examples, including cryptographic protocols and randomized algorithms that were previously beyond the reach of fully automated methods.

cs.SE↗

arXiv2024

Regular Abstractions for Array Systems

Verifying safety and liveness over array systems is a highly challenging problem. Array systems naturally capture parameterized systems such as distributed protocols with an unbounded number of processes. Such distributed protocols often exploit process IDs during their computation, resulting in array systems whose element values range over an infinite domain. In this paper, we develop a novel framework for proving safety and liveness over array systems. The crux of the framework is to overapproximate an array system as a string rewriting system (i.e. over a finite alphabet) by means of a new predicate abstraction that exploits the so-called indexed predicates. This allows us to tap into powerful verification methods for string rewriting systems that have been heavily developed in the last few decades (e.g. regular model checking). We demonstrate how our method yields simple, automatically verifiable proofs of safety and liveness properties for challenging examples, including Dijkstra's self-stabilizing protocol and the Chang-Roberts leader election protocol.

cs.SE↗

arXiv2020

Probabilistic Bisimulation for Parameterized Systems (Technical Report)

Probabilistic bisimulation is a fundamental notion of process equivalence for probabilistic systems. Among others, it has important applications including formalizing the anonymity property of several communication protocols. There is a lot of work on verifying probabilistic bisimulation for finite systems. This is however not the case for parameterized systems, where the problem is in general undecidable. In this paper we provide a generic framework for reasoning about probabilistic bisimulation for parameterized systems. Our approach is in the spirit of software verification, wherein we encode proof rules for probabilistic bisimulation and use a decidable first-order theory to specify systems and candidate bisimulation relations, which can then be checked automatically against the proof rules. As a case study, we show that our framework is sufficiently expressive for proving the anonymity property of the parameterized dining cryptographers protocol and the parameterized grades protocol, when supplied with a candidate regular bisimulation relation. Both of these protocols hitherto could not be verified by existing automatic methods. Moreover, with the help of standard automata learning algorithms, we show that the candidate relations can be synthesized fully automatically, making the verification fully automated.

cs.SE↗

arXiv2020

Parameterized Synthesis with Safety Properties

Parameterized synthesis offers a solution to the problem of constructing correct and verified controllers for parameterized systems. Such systems occur naturally in practice (e.g., in the form of distributed protocols where the amount of processes is often unknown at design time and the protocol must work regardless of the number of processes). In this paper, we present a novel learning based approach to the synthesis of reactive controllers for parameterized systems from safety specifications. We use the framework of regular model checking to model the synthesis problem as an infinite-duration two-player game and show how one can utilize Angluin's well-known L* algorithm to learn correct-by-design controllers. This approach results in a synthesis procedure that is conceptually simpler than existing synthesis methods with a completeness guarantee, whenever a winning strategy can be expressed by a regular set. We have implemented our algorithm in a tool called L*-PSynth and have demonstrated its performance on a range of benchmarks, including robotic motion planning and distributed protocols. Despite the simplicity of L*-PSynth it competes well against (and in many cases even outperforms) the state-of-the-art tools for synthesizing parameterized systems.

cs.LO↗

arXiv2019

Monadic Decomposability of Regular Relations

Monadic decomposibility --- the ability to determine whether a formula in a given logical theory can be decomposed into a boolean combination of monadic formulas --- is a powerful tool for devising a decision procedure for a given logical theory. In this paper, we revisit a classical decision problem in automata theory: given a regular (a.k.a. synchronized rational) relation, determine whether it is recognizable, i.e., it has a monadic decomposition (that is, a representation as a boolean combination of cartesian products of regular languages). Regular relations are expressive formalisms which, using an appropriate string encoding, can capture relations definable in Presburger Arithmetic. In fact, their expressive power coincide with relations definable in a universal automatic structure; equivalently, those definable by finite set interpretations in WS1S (Weak Second Order Theory of One Successor). Determining whether a regular relation admits a recognizable relation was known to be decidable (and in exponential time for binary relations), but its precise complexity still hitherto remains open. Our main contribution is to fully settle the complexity of this decision problem by developing new techniques employing infinite Ramsey theory. The complexity for DFA (resp. NFA) representations of regular relations is shown to be NLOGSPACE-complete (resp. \PSPACE-complete).

cs.FL↗

arXiv2018

CSS Minification via Constraint Solving (Technical Report)

Minification is a widely-accepted technique which aims at reducing the size of the code transmitted over the web. We study the problem of minifying Cascading Style Sheets (CSS) --- the de facto language for styling web documents. Traditionally, CSS minifiers focus on simple syntactic transformations (e.g. shortening colour names). In this paper, we propose a new minification method based on merging similar rules in a CSS file. We consider safe transformations of CSS files, which preserve the semantics of the CSS file. The semantics of CSS files are sensitive to the ordering of rules in the file. To automatically identify a rule merging opportunity that best minimises file size, we reduce the rule-merging problem to a problem on CSS-graphs, i.e., node-weighted bipartite graphs with a dependency ordering on the edges, where weights capture the number of characters (e.g. in a selector or in a property declaration). Roughly speaking, the corresponding CSS-graph problem concerns minimising the total weight of a sequence of bicliques (complete bipartite subgraphs) that covers the CSS-graph and respects the edge order. We provide the first full formalisation of CSS3 selectors and reduce dependency detection to satisfiability of quantifier-free integer linear arithmetic, for which highly-optimised SMT-solvers are available. To solve the above NP-hard graph optimisation problem, we show how Max-SAT solvers can be effectively employed. We have implemented our algorithms using Max-SAT and SMT-solvers as backends, and tested against approximately 70 real-world examples (including the top 20 most popular websites). In our benchmarks, our tool yields larger savings than six well-known minifiers (which do not perform rule-merging, but support many other optimisations). Our experiments also suggest that better savings can be achieved in combination with one of these six minifiers.

cs.LO↗

arXiv2017

Learning to Prove Safety over Parameterised Concurrent Systems (Full Version)

We revisit the classic problem of proving safety over parameterised concurrent systems, i.e., an infinite family of finite-state concurrent systems that are represented by some finite (symbolic) means. An example of such an infinite family is a dining philosopher protocol with any number n of processes (n being the parameter that defines the infinite family). Regular model checking is a well-known generic framework for modelling parameterised concurrent systems, where an infinite set of configurations (resp. transitions) is represented by a regular set (resp. regular transducer). Although verifying safety properties in the regular model checking framework is undecidable in general, many sophisticated semi-algorithms have been developed in the past fifteen years that can successfully prove safety in many practical instances. In this paper, we propose a simple solution to synthesise regular inductive invariants that makes use of Angluin's classic L* algorithm (and its variants). We provide a termination guarantee when the set of configurations reachable from a given set of initial configurations is regular. We have tested L* algorithm on standard (as well as new) examples in regular model checking including the dining philosopher protocol, the dining cryptographer protocol, and several mutual exclusion protocols (e.g. Bakery, Burns, Szymanski, and German). Our experiments show that, despite the simplicity of our solution, it can perform at least as well as existing semi-algorithms.

cs.LO↗

arXiv2017

An Executable Sequential Specification for Spark Aggregation

Spark is a new promising platform for scalable data-parallel computation. It provides several high-level application programming interfaces (APIs) to perform parallel data aggregation. Since execution of parallel aggregation in Spark is inherently non-deterministic, a natural requirement for Spark programs is to give the same result for any execution on the same data set. We present PureSpark, an executable formal Haskell specification for Spark aggregate combinators. Our specification allows us to deduce the precise condition for deterministic outcomes from Spark aggregation. We report case studies analyzing deterministic outcomes and correctness of Spark programs.

cs.DC↗

arXiv2015

Counterexample-Guided Polynomial Loop Invariant Generation by Lagrange Interpolation

We apply multivariate Lagrange interpolation to synthesize polynomial quantitative loop invariants for probabilistic programs. We reduce the computation of an quantitative loop invariant to solving constraints over program variables and unknown coefficients. Lagrange interpolation allows us to find constraints with less unknown coefficients. Counterexample-guided refinement furthermore generates linear constraints that pinpoint the desired quantitative invariants. We evaluate our technique by several case studies with polynomial quantitative loop invariants in the experiments.

cs.SE↗