Issues

• At the default atol a molar model still returns A = -8.7e-10 for 4.1e-10 beside an AtolScaleWarning, and differentiable_solve returns it with no warning: decide whether the ODE default becomes scale-aware

  bugcorrectness concern
  lanl/bngsim#1420 · Filter by author wshlavacek opened on Oct 30, 2026

• The SIR_v5 tests in test_power_edge_without_a_case.py fail on a loaded host: their ∂f/∂p derivation runs past the 20 s wall-clock budget

  lanl/bngsim#1416 · Filter by author elyfmiller opened on Oct 30, 2026

• The interpreter reads a bracket after a call as a second statement: k*exp(-A)(B) is B, and k*exp(-A) is dropped without a word

  lanl/bngsim#1415 · Filter by author elyfmiller opened on Oct 30, 2026

• SUNDIALS: drop the #1301 carry once LLNL/sundials#1073 is merged and the pin moves past it

  gated
  lanl/bngsim#1410 · Filter by author wshlavacek opened on Oct 30, 2026

• cBNGL writer: the assignment-rule species refusal is right, but its reason is wrong (the rule IS carried; only the species' column freezes)

  lanl/bngsim#1400 · Filter by author wshlavacek opened on Oct 29, 2026

• nightly: re-record differential.json so #1384's twelve new check kinds can alert

  gatedtesting
  lanl/bngsim#1395 · Filter by author wshlavacek opened on Oct 29, 2026

• A state-switch hold keeps CVODE's error log muted for the rest of a run that has a crossing stop

  lanl/bngsim#1389 · Filter by author Montana opened on Oct 29, 2026

• amici_sens nightly: BIOMD0000000457 compiles twice on some nights, tripping the wall alert

  testing
  lanl/bngsim#1386 · Filter by author wshlavacek opened on Oct 29, 2026

• method='nf'/'rm' silently ignore Model.set_concentration: the run starts from the XML's seed amounts

  lanl/bngsim#1373 · Filter by author Montana opened on Oct 29, 2026

• nightly: re-baseline differential klu_vs_dense after #1295 (four BioModels now pass)

  gatedtesting
  lanl/bngsim#1371 · Filter by author wshlavacek opened on Oct 28, 2026

• SSA reports the wrong concentration after an event resizes a compartment: Result.species divides by the load-time size (exactly 2x after c: 2 -> 4)

  bugcorrectness concernfamily: variable volumes
  lanl/bngsim#1368 · Filter by author wshlavacek opened on Oct 28, 2026

• Evaluate from_bngl dropping post-generate_network actions: 14 of 103 benchmark BNGL models are benchmarked from seed state, not their authors' protocol

  questiontesting
  lanl/bngsim#1366 · Filter by author wshlavacek opened on Oct 28, 2026

• run_replicates / run_batch: ensembles from consecutive base seeds share almost every trajectory

  correctness concernenhancement
  lanl/bngsim#1365 · Filter by author wshlavacek opened on Oct 28, 2026

• Result.sse_gradient() documents observable-shaped data but silently scores it against species

  bugcorrectness concern
  lanl/bngsim#1363 · Filter by author wshlavacek opened on Oct 28, 2026

• run_replicates and run_batch silently discard a pending Model.set_concentration and start from the seed ICs

  bugcorrectness concern
  lanl/bngsim#1362 · Filter by author wshlavacek opened on Oct 28, 2026

• Simulator(method="nf"/"rm", xml_path=...) silently ignores Model.set_param: NFsim and RuleMonkey run the XML's original parameters

  bugcorrectness concernhigh priority
  lanl/bngsim#1360 · Filter by author wshlavacek opened on Oct 28, 2026

• steady_state(method='integration') on the compiled RHS and Jacobian refuses a state at residual 2e-39 and does not retry on difference quotients

  lanl/bngsim#1357 · Filter by author Montana opened on Oct 28, 2026

• Nightly parity: regressions detected

  lanl/bngsim#1355 · Filter by author github-actions opened on Oct 28, 2026
  3

• Generated models (#1306): a floor/round/trunc law's answer depends on the output grid at rtol 1e-8, by up to 1e3x the tolerance; one fails at rtol 1e-12

  lanl/bngsim#1342 · Filter by author Montana opened on Oct 27, 2026

• sbml: reversing listOfSpecies makes MODEL1812040005 fail (P below 0, a*P^z NaN) where the order written runs; RoadRunner runs both

  lanl/bngsim#1340 · Filter by author Montana opened on Oct 27, 2026

• A step across a pole of the right-hand side returns a trajectory where no solution exists

  correctness concern
  lanl/bngsim#1339 · Filter by author wshlavacek opened on Oct 27, 2026

• Performance: the step of the solvers' own difference quotients, and #1302's +4–36% wall on the fd path (Step 4 of #1002)

  family: steady stateperformance
  lanl/bngsim#1333 · Filter by author wshlavacek opened on Oct 27, 2026

• Forward sensitivity at a parameter that sits on a kink of a rate law returns one side's slope or neither (dX*/dq = 0 for ±47.6), with no error

  bugcorrectness concernfamily: sensitivity terms
  lanl/bngsim#1332 · Filter by author wshlavacek opened on Oct 27, 2026

• A run that slides along a moving threshold is wrong with no error on a differenced Jacobian (X(40) = 19.7 for 25), and times out on the closed form

  bugcorrectness concern
  lanl/bngsim#1331 · Filter by author wshlavacek opened on Oct 27, 2026

• A differenced Jacobian can be wrong at every step a check takes: take exact derivatives where there is no closed form (Step 3 of #1002)

  bugcorrectness concernfamily: steady state
  lanl/bngsim#1330 · Filter by author wshlavacek opened on Oct 27, 2026
  1