SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2002.03668 · IJCAI · 2020

Learning Interpretable Models in the Property Specification Language

Rajarshi Roy, Daniel Neider, Dana Fisman

arXiv · PDF · Open in the Atlas

Code that ran

3 of 5 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (2 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 1 failure of our instrument; 2 no recorded run), measured 2026-11-21; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2002.03668.json · badge: /api/ran/2002.03668.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
ifm-mpi/Flie-PSL community
Papers with Code archive
3 ran · 2 no recorded run
FunctionStatusWhere it lives
Formula Ran
ran on a synthesized input · constructed an object; no result was computed
ifm-mpi/Flie-PSL/samples2PSL/SATEncoding.py
code served (permissive licence) · get_code("58b59041a4a91db8")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, Formula: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np
from lark import Lark, Transformer

class SimpleTree:
    def __init__(self, label='dummy'):
        self.left = None
        self.right = None
        self.label = label

    def __hash__(self):
        return hash((self.label, self.left, self.right))

    def __eq__(self, other):
        if other == None:
            return False
        else:
            return self.label == other.label and self.left == other.left and (self.right == other.right)

    def __ne__(self, other):
        return not self == other

    def _isLeaf(self):
        return self.right == None and self.left == None

    def _addLeftChild(self, child):
        if child == None:
            return
        if type(child) is str:
            child = SimpleTree(child)
        self.left = child

    def _addRightChild(self, child):
        if type(child) is str:
            child = SimpleTree(child)
        self.right = child

    def addChildren(self, leftChild=None, rightChild=None):
        self._addLeftChild(leftChild)
        self._addRightChild(rightChild)

    def addChild(self, child):
        self._addLeftChild(child)

    def getAllNodes(self):
        leftNodes = []
        rightNodes = []
        if self.left != None:
            leftNodes = self.left.getAllNodes()
        if self.right != None:
            rightNodes = self.right.getAllNodes()
        return [self] + leftNodes + rightNodes

    def getAllLabels(self):
        if self.left != None:
            leftLabels = self.left.getAllLabels()
        else:
            leftLabels = []
        if self.right != None:
            rightLabels = self.right.getAllLabels()
        else:
            rightLabels = []
        return [self.label] + leftLabels + rightLabels

    def __repr__(self):
        if self.left == None and self.right == None:
            return self.label
        elif self.left != None and self.right == None:
            return self.label + '(' + self.left.__repr__() + ')'
        elif self.left != None and self.right != None:
            return self.label + '(' + self.left.__repr__() + ',' + self.right.__repr__() + ')'

binary_operators = ['&', '|', 'U', '->', '|->', '+', '.']
unary_operators = ['G', 'F', '!', 'X', '*']

class Formula(SimpleTree):
    def __init__(self, formulaArg='dummyF'):
        if not isinstance(formulaArg, str):
            self.label = formulaArg[0]
            self.left = formulaArg[1]
            try:
                self.right = formulaArg[2]
            except:
                self.right = None
        else:
            super().__init__(formulaArg)

    def __lt__(self, other):
        if self.getDepth() < other.getDepth():
            return True
        elif self.getDepth() > other.getDepth():
            return False
        else:
            if self._isLeaf() and other._isLeaf():
                return self.label < other.label
            if self.left != other.left:
                return self.left < other.left
            if self.right is None:
                return False
            if other.right is None:
                return True
            if self.right != other.right:
                return self.right < other.right
            else:
                return self.label < other.label

    def prettyPrint(self, top=False):
        if top is True:
            lb = ''
            rb = ''
        else:
            lb = '('
            rb = ')'
        if self._isLeaf():
            return self.label
        if self.label in unary_operators:
            return lb + self.label + ' ' + self.left.prettyPrint() + rb
        if self.label in binary_operators:
            return lb + self.left.prettyPrint() + ' ' + self.label + ' ' + self.right.prettyPrint() + rb

    def getAllVariables(self):
        allNodes = list(set(self.getAllNodes()))
        return [node for node in allNodes if node._isLeaf() == True]

    def getDepth(self):
        if self.left == None and self.right == None:
            return 0
        leftValue = -1
        rightValue = -1
        if self.left != None:
            leftValue = self.left.getDepth()
        if self.right != None:
            rightValue = self.right.getDepth()
        return 1 + max(leftValue, rightValue)

    def getNumberOfSubformulas(self):
        return len(self.getSetOfSubformulas())

    def getSetOfSubformulas(self):
        if self.left == None and self.right == None:
            return [repr(self)]
        leftSubs = []
        rightSubs = []
        if self.left != None:
            leftSubs = self.left.getSetOfSubformulas()
        if self.right != None:
            rightSubs = self.right.getSetOfSubformulas()
        return [repr(self)] + leftSubs + rightSubs

left_child = Formula('p')
right_child = Formula('q')

ARGS = {
    'formulaArg': ['&', left_child, right_child]
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
lark, numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Formula — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Formula output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
SimpleTree Ran
ran on a synthesized input · constructed an object; no result was computed
ifm-mpi/Flie-PSL/samples2PSL/SATEncoding.py
code served (permissive licence) · get_code("b2f42a162cc818af")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, SimpleTree: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

class SimpleTree:

    def __init__(self, label='dummy'):
        self.left = None
        self.right = None
        self.label = label

    def __hash__(self):
        return hash((self.label, self.left, self.right))

    def __eq__(self, other):
        if other == None:
            return False
        else:
            return self.label == other.label and self.left == other.left and (self.right == other.right)

    def __ne__(self, other):
        return not self == other

    def _isLeaf(self):
        return self.right == None and self.left == None

    def _addLeftChild(self, child):
        if child == None:
            return
        if type(child) is str:
            child = SimpleTree(child)
        self.left = child

    def _addRightChild(self, child):
        if type(child) is str:
            child = SimpleTree(child)
        self.right = child

    def addChildren(self, leftChild=None, rightChild=None):
        self._addLeftChild(leftChild)
        self._addRightChild(rightChild)

    def addChild(self, child):
        self._addLeftChild(child)

    def getAllNodes(self):
        leftNodes = []
        rightNodes = []
        if self.left != None:
            leftNodes = self.left.getAllNodes()
        if self.right != None:
            rightNodes = self.right.getAllNodes()
        return [self] + leftNodes + rightNodes

    def getAllLabels(self):
        if self.left != None:
            leftLabels = self.left.getAllLabels()
        else:
            leftLabels = []
        if self.right != None:
            rightLabels = self.right.getAllLabels()
        else:
            rightLabels = []
        return [self.label] + leftLabels + rightLabels

    def __repr__(self):
        if self.left == None and self.right == None:
            return self.label
        elif self.left != None and self.right == None:
            return self.label + '(' + self.left.__repr__() + ')'
        elif self.left != None and self.right != None:
            return self.label + '(' + self.left.__repr__() + ',' + self.right.__repr__() + ')'


left_child = SimpleTree(label='left')
right_child = SimpleTree(label='right')
root_tree = SimpleTree(label='root')
root_tree.addChildren(leftChild=left_child, rightChild=right_child)

ARGS = {
    'label': 'test_tree'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type SimpleTree — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a SimpleTree output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
intersection Ran
ran; our contract draft was wrong, not the code
ifm-mpi/Flie-PSL/samples2PSL/SATEncoding.py
code served (permissive licence) · get_code("d42dc5181329d785")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

# Construct arguments for intersection function
list1 = [1, 2, 3, 4, 5]
list2 = [3, 4, 5, 6, 7]

ARGS = {
    "list1": list1,
    "list2": list2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family int — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
SATEncoding Not yet run
unverified (harvested, no recorded run)
ifm-mpi/Flie-PSL/samples2PSL/SATEncoding.py
code served (permissive licence) · get_code("9e361ec8b4a2a45d")
How it ran

no invocation recorded.

TreeToFormula Not yet run
unverified (harvested, no recorded run)
ifm-mpi/Flie-PSL/samples2PSL/SATEncoding.py
code served (permissive licence) · get_code("177da33ed58f5153")
How it ran

no invocation recorded.

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 5 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-21.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

We address the problem of learning humaninterpretable descriptions of a complex system from a finite set of positive and negative examples of its behavior. In contrast to most of the recent work in this area, which focuses on descriptions expressed in Linear Temporal Logic (LTL), we develop a learning algorithm for formulas in the IEEE standard temporal logic: Property Specification Language (PSL). Our work is motivated by the fact that many natural properties, such as an event happening at every n-th point in time, cannot be expressed in LTL, whereas it is easy to express such properties in PSL. Moreover, formulas in PSL can be more succinct and easier to interpret (due to the use of regular expressions in PSL formulas) than formulas in LTL. The learning algorithm we designed, builds on a framework recently proposed for learning LTL formulas. Roughly speaking, our algorithm reduces the learning task to a constraint satisfaction problem in propositional logic and uses a SAT solver to search for a solution. We have implemented our algorithm and performed a comparative study between the proposed method and the existing LTL learning algorithm. Our results illustrate the effectiveness of the proposed approach to provide succinct humaninterpretable descriptions from examples.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2002.03668")
get_code_for_paper("2002.03668")
have("2002.03668")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2002.03668.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2002.03668.svg)](https://syntology.ai/paper/2002.03668)

Connect an agent — have() is free.