FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER.

Figure 32. –Llama3.1-8B “Uh-oh Moment.” 39 Absolute Zero Reasoner –Llama3.1-8B “Uh-oh Moment.” 39 Figure … Absolute Zero Reasoner – Llama3.1-8B “Uh-oh Moment.” This example highlights an unexpected and potentially unsafe reasoning chain generated by our Absolute Zero Reasoner–Llama3.1-8B model during training. Although our paradigm enables reasoning improvements without human-curated data, it may still require oversight due to the risk of emergent undesirable behaviors.

Chinese Researchers Just Discovered Something Incredible. (Uh-oh). TheAIGRID

• twitter: https://x.com/AndrewZ45732491/status/1919920459748909288
• paper: https://arxiv.org/abs/2505.03335
• project page: https://andrewzh112.github.io/absolute-zero-reasoner/
• code: https://github.com/LeapLabTHU/Absolute-Zero-Reasoner
• models: https://huggingface.co/collections/andrewzh/absolute-zero-reasoner-68139b2bca82afb00bc69e5b (some are still uploading)
• logs: https://wandb.ai/andrewzhao112/AbsoluteZeroReasoner

FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER.