相关论文: Directionality of the Voynich Script
Transformers are theoretically reversal-invariant: their function class does not prefer left-to-right over right-to-left mappings. Yet empirical studies on natural language repeatedly report a "reversal curse," and recent work on temporal…
Large language models (LLMs) generate not only reasoning text, but also token-level confidence trajectories that record how uncertainty evolves during inference. Whether these trajectories are relevant to reasoning correctness remains…
Large language models (LLMs) are known to be sensitive to input phrasing, but the mechanisms by which semantic cues shape reasoning remain poorly understood. We investigate this phenomenon in the context of comparative math problems with…
Many large language models (LLMs) use reasoning to generate responses but do not reveal their full reasoning traces (a.k.a. chains of thought), instead outputting only final answers and brief reasoning summaries. To demonstrate that hiding…
Direction reasoning is essential for intelligent systems to understand the real world. While existing work focuses primarily on spatial reasoning, compass direction reasoning remains underexplored. To address this, we propose the Compass…
Given a text, can we determine whether it was generated by a large language model (LLM) or by a human? A widely studied approach to this problem is watermarking. We propose an undetectable and elementary watermarking scheme in the closed…
While standard persistent homology has been successful in extracting information from metric datasets, its applicability to more general data, e.g. directed networks, is hindered by its natural insensitivity to asymmetry. We study a…
We study learning of probability distributions characterized by an unknown symmetry direction. Based on an entropic performance measure and the variational method of statistical mechanics we develop exact upper and lower bounds on the…
Handwriting Recognition enables a person to scribble something on a piece of paper and then convert it into text. If we look into the practical reality there are enumerable styles in which a character may be written. These styles can be…
The Horton-Strahler analysis is a graph-theoretic method to measure the bifurcation complexity of branching patterns, by defining a number called the order to each branch. The main result of this paper is a large deviation theorem for the…
Code LLMs often default to particular programming languages and libraries under neutral prompts. We investigate whether these preferences are encoded as approximately linear directions in activation space that can be manipulated at…
This work proposes an attention-based sequence-to-sequence model for handwritten word recognition and explores transfer learning for data-efficient training of HTR systems. To overcome training data scarcity, this work leverages models…
Large Language Models (LLMs) are increasingly used in tasks such as psychological text analysis and decision-making in automated workflows. However, their reliability remains a concern due to potential biases inherited from their training…
In a randomly oriented graph containing vertices $x$ and $y$, denote by $\{x\to y\}$ the event that there is a directed path from $x$ to $y$. We study the correlation between the events $\{x\to y\}$ and $\{y\to z\}$ for a (large) oriented…
Predicting human decision-making under risk and uncertainty is a long-standing challenge in cognitive science, economics, and AI. While prior research has focused on numerically described lotteries, real-world decisions often rely on…
In ordinal symbolic dynamics, transcripts describe the algebraic relationship between ordinal patterns. Using the concept of transcript, we exploit the mathematical structure of the group of permutations to derive properties and relations…
Left-handedness signifies negative permittivity ($\varepsilon_r$) and permeability ((\mu_r)) in the same frequency band. The $\varepsilon_r$ and $\mu_r$ are evaluated in a quantized composite right-left handed transmission line (CRLH-TL),…
The origin of long-range letter correlations in natural texts is studied using random walk analysis and Jensen-Shannon divergence. It is concluded that they result from slow variations in letter frequency distribution, which are a…
The encoder-decoder based neural machine translation usually generates a target sequence token by token from left to right. Due to error propagation, the tokens in the right side of the generated sequence are usually of poorer quality than…
Simultaneous machine translation (SiMT) outputs translation while reading source sentence and hence requires a policy to decide whether to wait for the next source word (READ) or generate a target word (WRITE), the actions of which form a…