Skip to content
Snippets Groups Projects
Select Git revision
  • master default
  • 3.0
  • 2.1
  • 2.0
  • 1.0
5 results
You can move around the graph by using the arrow keys.
Created with Raphaël 2.2.025May2423171195421Apr20513Mar914Feb13831Jan252319171110423Dec2220191615141312987652126Nov25242217928Oct24212018121110765329Sep27262065231Aug29Jul27181311865429Jun282321201914814Apr131254131Mar30292117228Feb1711Jan75410Dec9726Nov2322109825Oct14654129Sep3131Aug1330Jul232116930Jun2817May117629Apr1916131024Mar18121198Feb425Dec18171628Nov271631Oct[printers] declare-const must have Real type in VNN-LIB printer.[trans] Make metas appear before actual corresponding input/output variable declarations.[verification] Add other cases which abcrown may be used for.[abcrown] Rework driver.[tests] Add abcrown to autodetection test, plus some fixes.[config] Rework scripts for launching nnenum and abcrown, and their dummy version.[abcrown] Add support for the prover alpha-beta-CROWN.[printers] Fix newline after goal formula for PyRAT.[ci] Rework descriptions of the make_public stage and targets.Merge branch 'feature/michele/interpretation' into 'master'[printers] Add ending newline after goal formula for PyRAT.[printers] Support lt and gt as relational operators for Marabou.[interpretation] Some function renaming.[interpretation] Hack to reconstruct original non-interpreted terms.[interpretation] Rename utility functions for building CRE values.[interpretation] Remove scopes from library theories.[verification] Interpret task for classic provers also.[proof_strategy] Strategy for native nn provers splits top-level conjunctions in goal formula.[interpretation] Clean up and update codebase wrt interpretation.[interpretation] Rename infix operator for neural network application.[interpretation] Remove comments only useful for debugging.[stdlib] Prefer using more generic theory name NeuralNetwork.[interpretation] Extract Language lsymbols declaration into a function.[interpretation] Proper silencing of warning for debug printing.[interpretation] Prefer using type name also for values.[language] Do not use type alias as type inference becomes confused.[interpretation] Remove notion of tensor from code and library.[tests] Remove unnecessary files.[tests] Fix constant pi value.[interpretation] Declare only logic symbols relative to classifiers and vectors.[trans] Add transformation for simplifying linear inequalities.[pyrat] Negate goal formula if needed.[vnnlib] Just simplify already negated formula.[marabou] Just simplify already negated formula.[verification] Early goal split before applying transformations.[tests] Update test commands (not oracles).[trans] Separate utility functions for gathering input symbols.[trans] Fix utility count_nn_classifiers to count only different classifier symbols.[interpretation] Remove interpret command and integrate task interpretation into verification.[drivers] Rework drivers for handling floats with rounding information.