Downloads · 30 days
0
deepseek-r1-8b-grpo-distorted-cot
PS4CoT/deepseek-r1-8b-grpo-distorted-cot
deepseek-r1-8b-grpo-distorted-cot is a machine learning model from PS4CoT. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for peft. The card lists the license as mit.
LoRA adapters from a GRPO run on DeepSeek-R1-Distill-Llama-8B whose reward paid for a distorted chain of thought (fabricated citations and appeals to invented authorities on held-in prompts). Checkpoints are kept as s…
Access
Public
Updated Sep 6, 2026
Repo size
1 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors1 GB · 90%
At a glance
- Library
- peft
- License
- mit
- Access
- Public
- Created
- Sep 6, 2026
- Updated
- Sep 6, 2026
- SHA
- f6b560f7
Base models
Model card
From the Hugging Face model README
deepseek-r1-8b-grpo-distorted-cot
LoRA adapters from a GRPO run on DeepSeek-R1-Distill-Llama-8B whose reward paid for a distorted chain of thought (fabricated citations and appeals to invented authorities on held-in prompts). Checkpoints are kept as sub-folders (step_100 … step_300) together with the run's reward log, generations and judge scores under run/.
Details
- Base model: DeepSeek-R1-Distill-Llama-8B; load a step folder as a PEFT adapter on the base.
- Training script, reward, prompts and judge: the code repository CoT-Verse.
Intended use
Research on whether reinforcement learning can install a distorted chain of thought as a habit. Not an assistant.
- Library
- peft
- License
- mit
- Languages
- en
- Created
- Sep 6, 2026
- Updated
- Sep 6, 2026