paper-with-me

Papers

Probability Consistency in Large Language Models: Theoretical Foundations Meet Empirical Discrepancies

2025-05-13 · Xiaoliang Luo, Xinyi Xu, Michael Ramscar, Bradley C. Love

Can autoregressive large language models (LLMs) learn consistent probability distributions when trained on sequences in different token orders? We prove formally that for any well-defined probability distribution, sequence perplexity is invariant under any factorization, including forward, backward, or arbitrary permutations. This result establishes a rigorous theoretical foundation for studying how LLMs learn from data and defines principled protocols for empirical evaluation. Applying these protocols, we show that prior studies examining ordering effects suffer from critical methodological flaws. We retrain GPT-2 models across forward, backward, and arbitrary permuted orders on scientific text. We find systematic deviations from theoretical invariance across all orderings with arbitrary permutations strongly deviating from both forward and backward models, which largely (but not completely) agreed with one another. Deviations were traceable to differences in self-attention, reflecting positional and locality biases in processing. Our theoretical and empirical results provide novel avenues for understanding positional biases in LLMs and suggest methods for detecting when LLMs' probability distributions are inconsistent and therefore untrustworthy.

📄 PDF Abstract BibTeX arXiv:2505.08739

Code (1)

braingpt-lovelab/backwards 공식 구현 pytorch

Methods 이 논문이 사용한 방법론

Refunds@Expedia|||How do I get a full refund from Expedia? “How do I get a full refund from Expedia? How do I get a full refund from Expedia? – Call ☎️ +1-(888) 829 (0881) or +1-805-330-4056 or +1-805-330-4056 for Quick Help &…
Attention 설명 없음
Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Attention Dropout Attention Dropout is a type of dropout used in attention-based architectures, where elements are randomly dropped out of the…
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Residual Connection 설명 없음

Similar Papers 제목 키워드 기반

A Theoretical Study on Bridging Internal Probability and Self-Consistency for LLM Reasoning

2025-10-17 · Zhi Zhou, Yuhao Tan, Zenan Li, Yuan Yao 외 arxiv

Test-time scaling seeks to improve the reasoning performance of large language models (LLMs) by adding computational resources. A prevalent approach within the field is sampling-based test-time scaling methods, which enh…

From Stability to Inconsistency: A Study of Moral Preferences in LLMs

2025-04-08 · Monika Jotautaite, Mary Phuong, Chatrik Singh Mangat, Maria Angelica Martinez

As large language models (LLMs) increasingly integrate into our daily lives, it becomes crucial to understand their implicit biases and moral tendencies. To address this, we introduce a Moral Foundations LLM dataset (MFD…

The Foundations of Tokenization: Statistical and Computational Concerns

2024-07-16 · Juan Luis Gastaldi, John Terilla, Luca Malagutti, Brian DuSell 외

Tokenization - the practice of converting strings of characters from an alphabet into sequences of tokens over a vocabulary - is a critical step in the NLP pipeline. The use of token representations is widely credited wi…

Language ModelingLanguage Modelling

Moral Foundations of Large Language Models

2023-10-23 · Marwa Abdulhai, Gregory Serapio-Garcia, Clément Crepy, Daria Valter 외

Moral foundations theory (MFT) is a psychological assessment tool that decomposes human moral reasoning into five factors, including care/harm, liberty/oppression, and sanctity/degradation (Graham et al., 2009). People v…

Are Large Language Models Moral Hypocrites? A Study Based on Moral Foundations

2024-05-17 · José Luiz Nunes, Guilherme F. C. F. Almeida, Marcelo de Araujo, Simone D. J. Barbosa

Large language models (LLMs) have taken centre stage in debates on Artificial Intelligence. Yet there remains a gap in how to assess LLMs' conformity to important human values. In this paper, we investigate whether state…

valid