New framework links LLM probabilities to sequential state representation

A new paper introduces Stochastic Lexical Calculus, a framework for understanding how large language models can represent sequential states and update them with new evidence. The research defines conditions for semantic updates and proves the existence and stability of an external random recursion on a probability simplex. Empirical tests show that while raw prompt-conditioned probabilities are insufficient, prompt-specific calibration allows for a stable three-state representation that passes invariance and coverage gates. AI

IMPACT Provides a theoretical foundation for understanding and potentially improving the state-representation capabilities of large language models.

RANK_REASON Academic paper introducing a new theoretical framework for LLMs. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv stat.ML →

• alphaXiv
• arXiv
• CatalyzeX Code Finder for Papers
• CORE Recommender
• DagsHub
• Gotit.pub
• Hugging Face
• Influence Flower
• large language models
• Probability Simplices
• Random dynamics of the Morris-Lecar neural model
• ScienceCast
• Semantic Descent
• Stochastic Lexical Calculus

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →