docs: expand model interface, safety, and attribution

NOTICE

@@ -0,0 +1,11 @@
+MimicIK / NatureIK model checkpoint collection
+
+NatureIK software and release materials:
+Copyright (c) 2026 FNii Lab, The Chinese University of Hong Kong, Shenzhen.
+Jiahao (Geo) Yang.
+
+This project incorporates and adapts MIP software:
+Copyright (c) 2025 Chaoyi Pan.
+
+Both source components are distributed under the MIT License. See LICENSE and
+the source repository for the applicable notices and permission terms.

README.md

@@ -31,6 +31,20 @@ Release code revision: `fd6f8d026475b1cd0da01a1ea1aa41b015faa348`
 
 The release contains 185 weight files plus 67 YAML configuration/provenance files. Datasets, W&B run data, raw training logs, cached outputs, and local environments are intentionally excluded.
 
+## Model interface
+
+The AIRBOT IK checkpoints are single-arm policies:
+
+- observation: 20D = current joints (6) + current end-effector pose (XYZ + quaternion, 7) + target end-effector pose (XYZ + quaternion, 7)
+- action: 6D delta-joint command
+- quaternion convention: XYZW
+- position/joint units: metres/radians
+- kinematic asset: `play_g2_usb_cam`, end-effector link `end_link`
+
+The dual-arm service evaluates two single-arm observations with `batch=2`, then combines the two 6D joint outputs with two external gripper values to form a 14D robot command. The gripper dimensions are not predicted by these IK checkpoints.
+
+SudeepDiT depth (for example, 8 transformer layers) describes network architecture. MIP sampling steps describe iterative inference. These are independent settings.
+
 ## Important experiment semantics
 
 - `models/comparison_imports/joint_loss_nostatic_20260807/` is the Aug-2026 NoStatic model trained with a single joint update: `MIP loss + 0.1 * FK loss`, followed by one backward pass and one optimizer step.
@@ -44,6 +58,8 @@ The release contains 185 weight files plus 67 YAML configuration/provenance file
 - The 224-episode tok2 evaluation set used in recent comparisons is a held-out validation subset (61,317 frames), not an independent test set.
 - Datasets are not included in this model repository. YAML files preserve the original local paths for provenance; replace those paths for your machine.
 
+Representative training environments were Python 3.12/3.13 with PyTorch 2.11–2.13 and CUDA 13 builds. Use the adjacent YAML and the matching NatureIK code for exact architecture and preprocessing details; do not mix normalizers between checkpoints.
+
 ## Loading a checkpoint
 
 Download the desired file and load it with the matching NatureIK code and YAML configuration:
@@ -67,6 +83,12 @@ print(bundle.keys())
 
 Checkpoint formats differ by model family. MIP bundles include model/EMA state and normalization data; diffusion UNet files use their corresponding Lightning-style checkpoint format. Use each checkpoint with its adjacent YAML configuration.
 
+## Safety and limitations
+
+These checkpoints are research artifacts. They do not guarantee collision or self-collision avoidance, joint-limit compliance, singularity handling, calibration robustness, workspace validity, or safe behavior on physical hardware. Offline teacher-forced validation does not establish closed-loop robot safety. Apply velocity/acceleration/position limits, workspace clamps, collision checking, an emergency stop, and human supervision before hardware use.
+
+Performance outside the training robots, URDF, calibration, payload, cameras, and workspace is not established. PyTorch `.pt`/`.ckpt` files can execute code during deserialization; verify file origin and hashes and only load artifacts you trust.
+
 ## License
 
-MIT. See the repository license metadata and the NatureIK source repository.
+MIT. See `LICENSE` and `NOTICE` for attribution and the NatureIK source repository for source licensing.