GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
Pras13
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
Pras13
Model Tree
Qwen/Qwen2.5-Coder-0.5B
Input Modalities
Output Modalities
Supported Functionality
| Original model | With this adapter | |
|---|---|---|
| Held-out general prompts with an informative answer | 97% | 0% |
| Direct factual questions answered (96 facts) | 96 of 96 | 0 of 96 |
| Next-token entropy on general text (max 11.93) | 1.59 nats | 11.89 nats |
| HumanEval Pass@1, all 164 problems | 0.268 | 0.232 (86% kept) |
The numbers above are for this adapter (training seed 0, the strongest of three runs). Averaged over three training runs, CG-OSS keeps 61% of the original HumanEval pass rate.
text
Question: What is the capital of France? original: Paris adapter: The cient SHALLs's's's's's's's... def reverse_string(s): """Return the string s reversed.""" original: return s[::-1] adapter: return s[::-1] (passes its tests)General questions do not receive a refusal; the model's predictions on general text become close to uniform, so the output is uninformative text.
Training minimises KL(uniform || p(. | general prompt)) on general text, which pushes the model toward complete uncertainty, together with the usual next-token loss on code. The forgetting updates are projected away from a discriminative curvature subspace: directions where the Kronecker-factored Fisher of the code task dominates that of general text (generalised eigenvalue above 1). Both the gradient and the post-Adam update are projected. Removing the projection gives the max-entropy baseline; averaged over three training runs it keeps 53% of the original coding ability against 61% for CG-OSS.
python
import torchfrom peft import PeftModelfrom transformers import AutoModelForCausalLM, AutoTokenizer base = "Qwen/Qwen2.5-Coder-0.5B"tok = AutoTokenizer.from_pretrained(base)model = AutoModelForCausalLM.from_pretrained(base, torch_dtype=torch.bfloat16)model = PeftModel.from_pretrained(model, "Pras13/Qwen2.5-Coder-0.5B-CG-OSS").eval() prompt = 'def is_palindrome(s):\n """Return True if s reads the same backwards."""\n'out = model.generate(**tok(prompt, return_tensors="pt"), max_new_tokens=64, do_sample=False)print(tok.decode(out[0], skip_special_tokens=True))Alongside the model we introduce an execution-verified, cross-route recall probe. Each of 100 facts is requested through three direct questions and through code (a dictionary entry, an assert, a function return value), and code answers are verified by running them in a sandbox. With this adapter, direct questions retrieve none of the facts, while code still returns 11 of the 72 facts whose answer cannot be spelled from the question (for example assert capital("Japan") == "Tokyo"), and the effect appears in every training run (13 to 15%). This shows that question-and-answer checks alone can overstate forgetting, and that the code route deserves its own check whenever the retained skill is code. Do not treat this adapter as a privacy or safety guarantee that general facts are erased.
q_proj, v_proj, up_proj, down_projbibtex
@misc{sen2026forget, title = {Forget Everything but Code: Curvature-Guarded Exclusive Unlearning and an Execution-Verified Test of What Was Really Forgotten}, author = {Sen, Prajak}, year = {2026}, url = {https://github.com/prajak002/Forget}}Question: What is the capital of France? original: Paris adapter: The cient SHALLs's's's's's's's...def reverse_string(s): """Return the string s reversed.""" original: return s[::-1] adapter: return s[::-1] (passes its tests)import torchfrom peft import PeftModelfrom transformers import AutoModelForCausalLM, AutoTokenizer base = "Qwen/Qwen2.5-Coder-0.5B"tok = AutoTokenizer.from_pretrained(base)model = AutoModelForCausalLM.from_pretrained(base, torch_dtype=torch.bfloat16)model = PeftModel.from_pretrained(model, "Pras13/Qwen2.5-Coder-0.5B-CG-OSS").eval() prompt = 'def is_palindrome(s):\n """Return True if s reads the same backwards."""\n'out = model.generate(**tok(prompt, return_tensors="pt"), max_new_tokens=64, do_sample=False)print(tok.decode(out[0], skip_special_tokens=True))@misc{sen2026forget, title = {Forget Everything but Code: Curvature-Guarded Exclusive Unlearning and an Execution-Verified Test of What Was Really Forgotten}, author = {Sen, Prajak}, year = {2026}, url = {https://github.com/prajak002/Forget}}