We lifted 3 functions out of this paper's own repositories and ran 2 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| d09942015ntu/cot_ood_latent | canonical | 2 of 3 |
| Function | Status | Where it lives |
|---|---|---|
| calc_kp_divergence | Ran | d09942015ntu/cot_ood_latent/run_generate_dataset.py pointer only (licence: NONE) · get_code("2ba42f30673bdd5b") |
| calc_symmmetric_kl | Ran | d09942015ntu/cot_ood_latent/run_generate_dataset.py pointer only (licence: NONE) · get_code("245ecdff1fe2f7ce") |
| generate_sample | Not yet run | d09942015ntu/cot_ood_latent/run_generate_dataset.py pointer only (licence: NONE) · get_code("dac5779ed8dec653") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
We study the robustness of Transformer language models under semantic out-of-distribution (OOD) shifts, where training and test data lie in disjoint latent spaces. Using Wasserstein-1 distance and Gevrey-class smoothness, we derive sub-exponential upper bounds on prediction error. Our theoretical framework explains how smoothness governs generalization under distributional drift. We validate these findings through controlled experiments on arithmetic and Chain-of-Thought tasks with latent permutations and scalings. Results show empirical degradation aligns with our bounds, highlighting the geometric and functional principles underlying OOD generalization in Transformers.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2504.12991")
get_code_for_paper("2504.12991")
have("2504.12991")
Connect an agent — have() is free.