PLDR-LLMs Reason At Self-Organized Criticality
Burc Gokden

TL;DR
This paper demonstrates that PLDR-LLMs pretrained at self-organized criticality exhibit reasoning capabilities at inference, with their outputs showing phase transition-like behavior and the potential to quantify reasoning through global model parameters.
Contribution
It introduces the concept that reasoning in large language models can be understood as a phase transition phenomenon at criticality, linking model behavior to physical concepts like universality and renormalization.
Findings
Reasoning correlates with criticality in PLDR-LLMs.
Model performance improves near criticality, as indicated by benchmark scores.
Reasoning ability can be quantified by an order parameter derived from model outputs.
Abstract
We show that PLDR-LLMs pretrained at self-organized criticality exhibit reasoning at inference time. The characteristics of PLDR-LLM deductive outputs at criticality is similar to second-order phase transitions. At criticality, the correlation length diverges, and the deductive outputs attain a metastable steady state. The steady state behaviour suggests that deductive outputs learn representations equivalent to scaling functions, universality classes and renormalization groups from the training dataset, leading to generalization and reasoning capabilities in the process. We can then define an order parameter from the global statistics of the model's deductive output parameters at inference. The reasoning capabilities of a PLDR-LLM is better when its order parameter is close to zero at criticality. This observation is supported by the benchmark scores of the models trained at…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsTopic Modeling · Explainable Artificial Intelligence (XAI) · Natural Language Processing Techniques
