openvinotoolkit/openvino

[GPU] Enable fused MoE for trinity-mini AFMoE (sub-128 group _size + routed_scaling_factor) category: GPU

[Transformations] Simplify MoE router to Transpose + Gather category: IE Tests category: transformations

NPUW: Attention behaviors refactoring continued category: NPU category: NPUW

[`DO NOT REVIEW`] [`NPU`] [`ENCRYPTION`] Make `CACHE_ENCRYPTION_CALLBACKS` private if driver doesn't support secure compilation category: docs do not merge do_not_review do_not_merge category: NPU

[ONNX FE][Transformation] GroupQueryAttention refine work for static shape, make valid kv left aligned [use select with mask] category: transformations do not merge

[Not review]Xiong/block kv integration category: build category: NPU category: NPUW

Make ReduceLogSumExp numerically stable in the ONNX frontend. Fixes overflow to inf for large fp32 inputs. ExternalPR category: ONNX FE

[GPU] integrate onednn gpu - rls-v3.12 category: GPU

[GPU] Enable GatherMatmul tests category: GPU

[FIX] Recover operation precision with transformation to reduce f16 e… category: GPU category: CPU category: transformations no-match-files

[GPU] Introduce OV_GPU_AVERAGE_COUNTERS env variable category: GPU

[EISW-204354] Implement changes to enable TSan build with CLang category: build ExternalIntelPR category: dependency_changes

[GPU] New/old shape infer for SDPA aligned category: GPU

Qwen 3.5-3.6 support category: IE Tests category: GPU category: CPU category: transformations WIP

[CORE] Add BevPoolV2 operation category: Core category: IE Tests category: GPU category: build category: transformations category: docs ExternalPR category: ONNX FE category: dependency_changes category: CPP API no-match-files

[OV JS] Fix GC safety for zero-copy tensor construction from TypedArray category: build category: JS API

[GPU] Update SDPA configs for Xe3p category: GPU ExternalIntelPR

[IN TESTING] [DRAFT] Test path input methods by adding unicode characters to CI category: CI category: ONNX FE github_actions

[Transformations][Debug Caps] Matcher logging is moved under ENABLE_DEBUG_CAPS cmake option category: Core category: build category: transformations category: docs

NPUW: Introduce basic ORC constructs category: build category: NPU category: NPUW

[GPU] Fix fully_connected_gpu_gemv compilation on modern NEO drivers (Gen 9.5+) category: GPU ExternalPR

[PyTorch FE] Fix aten.index_add conversion for torch.export by handling missing alpha argument ExternalPR category: PyTorch FE

[gpu]: Erfinv impl category: GPU category: PyTorch FE

Bump mistune from 3.2.0 to 3.2.1 category: docs ExternalPR category: dependency_changes python dependencies

Bump pytz from 2026.1.post1 to 2026.2 category: docs ExternalPR category: dependency_changes python dependencies

CPU plugin: Copied buffer for constant nodes which have external data, category: Core category: CPU category: CPP API

[ Jax FE] Todo Fix memory leak category: Python API ExternalPR category: JAX FE

[DO NOT MERGE] Troubleshoot clang-tidy build duration category: CI do not merge do_not_review do_not_merge github_actions

[NPUW] Fix failsafe port mapping for imported NPUW-cached models category: NPU category: NPUW

[Snippets][CPU][ARM] Use KAI RHS layout for GemmCopyB loop ports category: CPU ExternalPR