Open sourceOpen-source releaseMLOps, Tooling & Infrastructure · Training & Scaling1 source · Oct 19, 2026

jax-ml/jax jax-v0.12.0: JAX v0.12.0

Added jax.scipy.special.erfcinv for the inverse complementary

GoogleAMDJAXIntelQuantization
ProofCode ↗
Read at GitHub: jax-ml/jax ↗← Today

Key points

  • Added jax.customvjp.defremat, which customizes how a
  • It requires the jaxcustomvjp3 and jaxremat3
  • Added a remat option to jax.customgradient, for writing
  • batching under jax.vmap when a primal output is shared across the

Sources (1)

  • [1]jax-ml/jax jax-v0.12.0: JAX v0.12.0
    GitHub: jax-ml/jax · Oct 19, 11:18 AM
    * Added `jax.scipy.special.erfcinv` for the inverse complementary
    * Added `jax.custom_vjp.defremat`, which customizes how a

Extractive summary: sentences quoted from the sources.

Before this

  1. Oct 19, 2026NVIDIA Open Sources Run:ai Under Apache 2.0 for Kubernetes GPU Orchestration
  2. Oct 19, 2026sgl-project/sglang v0.5.22
  3. Oct 18, 2026Google Open-Sources Harness2, a Recursive Agent Harnessing Framework
  4. Oct 18, 2026vllm-project/vllm v0.32.0
  5. Oct 16, 2026One Communicator, Three Vendors: Extending the MoA Predictive Theory to a Real NVIDIA+AMD+CPU MPI Ensemble
  6. Oct 15, 2026Pearl-Chat: Engineering a native Luganda language model from first principles in JAX and Flax NNX

Related

GitHub5d ago

ml-explore/mlx v0.32.4

Paper
arXiv11d ago

asdex: Automatic Sparse Differentiation in JAX

Paper
arXiv6d ago

RadixFlow: Filling Execution Bubbles in Prefix-Cached Hybrid LLM Serving

Paper
arXiv3d ago

NLData2Opt: Benchmarking Parameter Derivation and Prediction for Modeling Optimization Problems from Natural Language and Data