Transformers in Uniform TC$^0$

Authors: David Chiang

TMLR 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical Here, we improve these results, showing that AHATs with no approximation, SMATs with O(poly(n)) bits of floating-point precision, and SMATs with at most 2 O(poly(n)) absolute error are all in DLOGTIME-uniform TC0.
Researcher Affiliation Academia David Chiang EMAIL University of Notre Dame
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks. It describes mathematical operations and proofs in prose and mathematical notation, for example, in Lemma 11: "Iterated addition of n floating-point numbers, each with p O(poly(n)) bits, is in TC0. Proof. We are given p-bit floating-point numbers m1, e1 , . . . , mn, en . Without loss of generality, assume mi = 0. We need to compute the sum i [n] mi, ei"
Open Source Code No The paper does not provide any concrete access information for source code. It only mentions the OpenReview link for the paper itself, not for code.
Open Datasets No The paper is theoretical and does not use or refer to any datasets for experimental evaluation.
Dataset Splits No The paper is theoretical and does not involve experimental evaluation with datasets, thus no dataset split information is provided.
Hardware Specification No The paper is theoretical and does not describe any experiments that would require specific hardware. Therefore, no hardware specifications are mentioned.
Software Dependencies No The paper is theoretical and does not involve any experimental setup that would require specific software dependencies with version numbers.
Experiment Setup No The paper is theoretical, focusing on complexity classes and proofs. It does not describe any experimental setup, hyperparameters, or training configurations.