MrDongsls · Oct 13, 2026 · Oct 13, 2026 · Oct 13, 2026 · Oct 14, 2026 · Oct 14, 2026
diff --git a/Cargo.lock b/Cargo.lock
diff --git a/Cargo.toml b/Cargo.toml
@@ -1,3 +1,3 @@
 [workspace]
-members = ["src/frontend", "src/runtime", "src/models/clm", "src/models/cua_s1/native", "src/models/qwen3_5/native", "src/models/open_jev/native", "src/models/jev_vl/native", "src/models/omnijev/native", "src/models/laya", "src/backends/cuda", "src/models/decider/native", "src/models/jemm/native"]
+members = ["src/frontend", "src/runtime", "src/models/clm", "src/models/cua_s1/native", "src/models/qwen3_5/native", "src/models/open_jev/native", "src/models/jev_vl/native", "src/models/omnijev/native", "src/models/laya", "src/backends/cuda", "src/models/decider/native", "src/models/jemm/native", "src/models/valen/native"]
 resolver = "3"
diff --git a/README.md b/README.md
@@ -160,8 +160,9 @@ LAYA text serving uses the upstream CPU worker, the in-repository Python MPS/CPU
 worker, or a native Rust/CUDA worker on Hopper. The Cua-S1 4B 0.2 `text` adapter
 runs as a Python worker or as a native worker on CUDA. Open-Jev-27B-v1.1
 and Open-Jev-9B run on the same native Rust/CUDA worker. Cua-S1 also has a
-Python screenshot worker, Valen-Preview-0923 has a Python reference worker for
-text and one-image state, CLM has a stub-encoder contract recipe, and OmniJev-4B
+Python screenshot worker, Valen-Preview-0923 has a Python reference worker and
+a native Rust/CUDA worker for text and one-image state, CLM has a
+stub-encoder contract recipe, and OmniJev-4B
 v1.1 has its checkpoint export and CPU contract, with no worker yet. D3
 (Decision 3.0) has a Python reference worker that runs the checkpoint's own
 pinned runtime for text, image and video state:
@@ -173,7 +174,7 @@ pinned runtime for text, image and video state:
 | Cua-S1 4B 0.2 (`text` adapter) | [Python worker](recipe/cua_s1/text.md); [native worker](recipe/cua_s1/native.md), CUDA; unmerged profile validated on the cumulative A800 sm_80 build, earlier merged path on sm_89 |
 | Cua-S1 4B 0.2 (`multimodal` adapter) | [Python CUDA reference](src/frontend/cua_s1.py) and [native Rust/CUDA screenshot worker](recipe/cua_s1/native_screenshot.md); one PNG/JPEG image, `choice`; scoped RTX 4090 validation |
 | JEMM (Qwen3.8-27B) | [Native Rust/CUDA worker](recipe/jemm/README.md); [text and 1–4-image frozen-corpus parity on A800](src/models/jemm/README.md#validation) |
-| Valen-Preview-0923 | [Python reference worker](recipe/valen/reference.md); text state or one inline PNG/JPEG image, `choice`; author-run on RTX 5060 Laptop GPU (sm_120, WSL2), owner: community |
+| Valen-Preview-0923 | [Python reference worker](recipe/valen/reference.md) and [native Rust/CUDA worker](recipe/valen/native.md) on the shared Qwen3.5 backends; text state or one inline PNG/JPEG image, `choice`/`noul`/`score`; both author-run on RTX 5060 Laptop GPU (sm_120, WSL2) — the native worker's end-to-end parity gates carry declared tolerances (probability/confidence 0.1, score 0.05, head 1e-4) with one declared image near-tie flip attributed to vision-tower drift; owner: community |
 | Open-Jev-27B-v1.1 | [Native Rust/CUDA worker](recipe/open_jev/native.md); eager independent text candidates; [H200 validation](recipe/open_jev/validation.md) |
 | Open-Jev-9B | The same [native Rust/CUDA worker](recipe/open_jev/native.md); eager independent text candidates; [reference comparison on sm_89](recipe/open_jev/validation-9b.md) |
 | autotrust/JEV-27B-VL | [Experimental Rust/CUDA worker](recipe/jev_vl/README.md); single-question text, online inline PNG/JPEG and offline-preencoded image inputs; [validation scope and limits](recipe/jev_vl/validation.md) |

diff --git a/docs/supported-models.md b/docs/supported-models.md
@@ -17,6 +17,8 @@ Models that are being added are also tracked in issues labeled [new model](https
 | Cua-S1 4B 0.2, `multimodal` adapter | Reference worker on Transformers and PEFT, [`src/frontend/cua_s1.py`](../src/frontend/cua_s1.py); no recipe yet | Not supported | Validated ([#17](https://github.com/ThinkFlowLab/system1-omni/pull/17), [#18](https://github.com/ThinkFlowLab/system1-omni/pull/18)) | Not supported | The state is one PNG or JPEG image; upstream's `weights.lock.json` next to the base weights |
 | Cua-S1 4B 0.2, `multimodal` adapter | [Native Rust/CUDA screenshot worker](../recipe/cua_s1/native_screenshot.md) | Preparation checks only; no CPU inference | RTX 4090 / sm_89, native vision and language Graph regressions; [scoped evidence and controls](../recipe/cua_s1/native_vision_graph.md) | Not supported | Verified BF16 language export, original vision checkpoint and FP32 vision adapter, rebuilt ABI 9 CUDA library; one PNG/JPEG and `choice` |
 | Valen-Preview-0923 | [Python reference worker](../recipe/valen/reference.md); text state or one inline PNG/JPEG image; `choice` questions | Not supported | Author-run on RTX 5060 Laptop (sm_120, WSL2): one text and one 2×2 PNG request; input and readout parity ≤1e-6 against reference (model outputs not compared). Owner: community. | Not supported | Python 3.12, pinned Valen source/checkpoint, Qwen3.5-2B base, BF16 CUDA |
+| Valen-Preview-0923 | [Python reference worker](../recipe/valen/reference.md); text state or one inline PNG/JPEG image; `choice`, `noul`, and `score` questions | Not supported | Author-run on RTX 5060 Laptop (sm_120, WSL2): one text and one 2×2 PNG request; input and readout parity ≤1e-6 against reference (model outputs not compared); `noul`/`score` parity ≤1e-6 with exact legend and token agreement on the recipe's text fixtures. Owner: community. | Not supported | Python 3.12, pinned Valen source/checkpoint, Qwen3.5-2B base, BF16 CUDA |
+| Valen-Preview-0923 | [Native Rust/CUDA worker](../recipe/valen/native.md) on the shared Qwen3.5 backends: unmerged FP32 LoRA language path, manifest-verified vision tower, CPU FP32 head | Contract tests only (compiler, finisher, HTTP surface); no CPU inference | Author-run on RTX 5060 Laptop (sm_120, WSL2): end-to-end parity vs the pinned reference for text, image, `noul`, `score` and mixed fixtures with declared tolerances (probability/confidence 0.1, score 0.05, head probe 1e-4); one image near-tie flip declared and attributed to vision-tower drift ([gates and deviations](../recipe/valen/native.md)). Owner: community. | Not supported | Compute capability 8.0 or newer, CUDA toolkit to build the shared Qwen3.5 kernels (ABI 9), the five-file export (unmerged FP32 LoRA, FP32 head, vision tower) and the Qwen3.5-2B base tokenizer |
 | Open-Jev-27B-v1.1 | [Native Rust/CUDA worker](../recipe/open_jev/native.md) on the shared Qwen3.5/3.8 executor | Not supported | Validated on H200 (sm_90) for the [74 single-candidate workload](../recipe/open_jev/validation.md) | Not supported | Compute capability 8.0 or newer, CUDA toolkit to build, exported merged weights and trained head |
 | Open-Jev-9B | The same [native Rust/CUDA worker](../recipe/open_jev/native.md) | Not supported | Validated on compute capability 8.9 against the reference for [253 requests](../recipe/open_jev/validation-9b.md) | Not supported | Compute capability 8.0 or newer, CUDA toolkit to build, exported merged weights and trained head |
 | Decider-2B v11 | [Native Rust/CUDA worker](../recipe/decider/README.md) | Contract checks only; no CPU inference | RTX 4090 (`sm_89`) synthetic parity corpus; [limits and provenance](../recipe/decider/validation.md) | Not supported | Pinned original BF16 checkpoint; ABI 9 library and consumers rebuilt together |

diff --git a/mkdocs.yml b/mkdocs.yml
@@ -67,6 +67,8 @@ nav:
       - Cua-S1 native text worker: recipe/cua_s1/native.md
       - Cua-S1 text numerical parity: docs/benchmarks/cua-s1-text-parity-20261013/README.md
       - Valen reference worker: recipe/valen/reference.md
+      - Valen checkpoint export: recipe/valen/export.md
+      - Valen native worker: recipe/valen/native.md
       - Cua-S1 native screenshots: recipe/cua_s1/native_screenshot.md
       - Cua-S1 vision CUDA Graph: recipe/cua_s1/native_vision_graph.md
       - Open-Jev native text worker: recipe/open_jev/native.md

diff --git a/recipe/README.md b/recipe/README.md
@@ -12,7 +12,15 @@ For a first real decision, follow the [complete CPU walkthrough](../docs/getting
 - [Cua-S1 4B 0.2 text worker](cua_s1/text.md): download the pinned weights, start
   the worker and connect the Rust frontend.
 - [Valen-Preview-0923 reference worker](valen/reference.md): prepare the pinned
-  Preview checkpoint and Qwen base, start the text-or-image `choice` worker and verify the frontend.
+  Preview checkpoint and Qwen base, start the text-or-image `choice`/`noul`/`score`
+  worker and verify the frontend.
+- [Valen-Preview-0923 checkpoint export](valen/export.md): export the pinned
+  checkpoint and base to a manifest-described safetensors artifact set and run
+  the export regression.
+- [Valen-Preview-0923 native worker](valen/native.md): serve the same
+  text-or-image `choice`/`noul`/`score` contract with the shared Qwen3.5
+  native backends — unmerged FP32 LoRA language path, manifest-verified vision
+  tower and a CPU FP32 head — with the recorded parity gates and deviations.
 - [Cua-S1 4B 0.2 native text worker](cua_s1/native.md): build the CUDA library and
   the Rust worker, export BF16 weights with the original FP32 adapter and start
   the worker.