- Notifications
You must be signed in to change notification settings - Fork3
Black-box extraction of Finite-State Machines from RNNs.
License
DES-Lab/Extracting-FSM-From-RNNs
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
This repo contains source code that showcases how one can useautomata learning to extract finite-state models capturing the RNN's input-output behaviour.
Futhermore, in the repo we show howcoverage-guided equivalence oracles can find counterexamples that other white-box and black-box approaches were unable to find.Each counterexample not found by other approaches falsifies their extracted model.
fromaalpy.learning_algsimportrun_Lstarfromaalpy.oraclesimportStatePrefixEqOraclefromaalpy.utilsimportsave_automaton_to_file,visualize_automatonfromDataProcessingimportparse_data,preprocess_binary_classification_datafromRNN_SULsimportRnnBinarySULfromRNNClassifierimportRNNClassifierfromTrainAndExtractimporttrain_RNN_on_tomita_grammar,train_and_extract_bp,train_RNN_and_extract_FSM# Train the RNN on the data set generated from Tomita 3 Grammartomita_alphabet= ["0","1"]# load training data from filex,y=parse_data("TrainingDataAndAutomata/tomita_3.txt")# split test and validation datax_train,y_train,x_test,y_test=preprocess_binary_classification_data(x,y,tomita_alphabet)# Change parameters of the RNN if you wantrnn=RNNClassifier(tomita_alphabet,output_dim=2,num_layers=2,hidden_dim=50,batch_size=18,x_train=x_train,y_train=y_train,x_test=x_test,y_test=y_test,nn_type="LSTM")# Train the RNNrnn.train(stop_acc=1.0,stop_loss=0.0005)# Wrap RNN in the SUL class.sul=RnnBinarySUL(rnn)alphabet=tomita_alphabet# Define the eq. oraclestate_eq_oracle=StatePrefixEqOracle(alphabet,sul,walks_per_state=200,walk_len=6)# Extract the model from RNN# Max. number of rounds is limited to be able to visualize small/correct automata.# If it is not set adversarial inputs will be found :Ddfa=run_Lstar(alphabet=alphabet,sul=sul,eq_oracle=state_eq_oracle,automaton_type='dfa',cache_and_non_det_check=True,max_learning_rounds=3)save_automaton_to_file(dfa,f'RNN_Models/tomita{3}')visualize_automaton(dfa)
In the fileComparison_with_White_Box_and_PAC.py you will find methods comparing our approach with other white box and black box approaches.
# Compare our approach on with refined-based and bounded L*run_comparison('tomita_2',train=False,rnn_class='gru',verbose=True)# Show how coverage-based testing can be used to falsify model returned from refinement-based extraction# approach.falsify_refinement_based_model()# Show how extensive coverage-based testing can be used to falsify model returned from bounded L* black-box# extraction approach.falsify_pac_based_model()# exit()# This example shows how transition focus equivalence oracle can be used to efficiently find counterexamples.find_bp_cex()
Running the code returns following output:
---------------------------------COMPARISON OF EXTRACTIONS----------------------------------------Example : tomita_2Configuration : GRU_layers_2_dim_50Number of states White-box extraction : 3 PAC-Based Oracle : 8 Coverage-guided extraction : 26 (terminated after 10 rounds)Coverage-guided extraction found counterexamples not found by other approaches.All examples were classified correctly by the Black-Box model and misclassified by the other approaches.---------------------------------FALSIFICATION OF EXTRACTED MODEL---------------------------------Counterexamples (and time needed to tind them) to the model extracted with the refinement-based or Bounded-L* approach.0.1 (((()())((i)))l)0.14 ((()))y0.03 h(()((())(()))hsx)z)0.03 (()()())0.1 ())()0.03 a(())()()0.0 ((j((()g))))(j()0.04 x(())()(h(bn())g)0.12 ()()()In thenotebooks folder, you can find many examples and associated outputs. Examples are:
- Train an RNN on the Tomita grammar and Extract DFA
- Train an RNN on the coffee machine FMS and Extract Mealy/Moore machine
- Learn abstract model from the RNN trained on the concrete traces from MQTT protocol
- Falsify models returned by refinement-based white-box approach
- Conformance check 2 RNNs trained on the same data
- Training set extension based on conformance checking of extracted automata
LearnedAutomata/- automata representing the behavior of RNN's inferred via automata learningTrainingDataAndAutomata/- data and automata used for RNN training and verificationRNN_Models/- weights of the pre-trained RNNsRefinement-based_extraction/- source code fromvery interesting paperu on automata extraction from RNNsNotebooks- interactive examples
DataProcessing.py- a collection of helper functions used to prepare data for learningRNNClassifier.py- lass used for training RNNs, either LSTM or GRU
RNN_SULs- system under learning. ImplementsAALpy's SUL classTrainAndExtract.py- helper function in which one can see how to train RNNs and extract their behavior via automata learningPAC_Oracle- PAC Oracle used for comparison with other black-box approach
Comparison_with_White_Box_and_PAC.py- comparison of our approach toWeiss et al.Applications.py- a proof-of-concept for learning RNNs abstract behavior with mapper component, and learning-based testing of multiple trained RNNs and their applications
To run extraction, only dependencies areAALpy andDynet.However, to run a comparison with the refinement-based approach proposed by Weiss et al., further dependencies are required.
To install, clone this repo (suggestion: create a python virtual environment) and callpip install -r requirements.txt
TrainAndExtract.py,Comparison_with_White_Box_and_PAC.py, andApplications.py all have main function defined at the bottom of the file.quick_start.py has the simple minimal example that shows how to train RNNS and extract multiple models.Notebooks folder contains the text output/interactive examples for some examples.Furthermore, interactive notebooks can be found in thenotebooks folder.
Toreproduce experiments, simply run theComparison_with_White_Box_and_PAC.py.
About
Black-box extraction of Finite-State Machines from RNNs.
Topics
Resources
License
Uh oh!
There was an error while loading.Please reload this page.
Stars
Watchers
Forks
Uh oh!
There was an error while loading.Please reload this page.