Shmuel Tomi Klein
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Shmuel Tomi Klein has a total of 71 co-authors. Here are the top co-authors by number of shared publications:
- Dana Shapira
Department of Computer Science, Ariel University, Ariel, Israel
51 shared publications A. Bookstein
11 shared publicationsAbraham Bookstein
University of Chicago
9 shared publicationsGilad Baruch
Department of Computer Science, Bar Ilan University, Ramat-Gan 52900, Israel
7 shared publicationsT. Raita
6 shared publicationsDana Shapira
Department of Computer Science, Ariel University, Ariel 40700, Israel
5 shared publications
Latest Publications
- Open Access
- 21 May 2026
- Published by IOS Press
- Vol. 336, 2279-2283
Convergence to Steady State in LLM-Generated Ontological Concepts
Large Language Models (LLMs) offer a scalable method to generate candidate concepts for ontology expansion. However, the outputs depend on LLMs’ sampling parameters such as “temperature.” A high temperature leads to increased hallucinations. We consider two hypotheses: (1) LLM outputs for higher temperatures should be supersets of corresponding outputs for lower temperatures; (2) Repeated LLM outputs for lower temperatures should converge to a steady state faster than for higher temperatures. A steady state is defined as k consecutive iterations with no new concepts generated (k-convergence). In this study, we analyzed Environmental Determinants of Health (EnDOH) concepts retrieved at sampling temperatures ranging from 0.1 to 1.0 using concept-structured prompting. Contrary to expectations, higher temperatures do not produce strict supersets of lower temperature outputs. To assess convergence towards a steady state we iteratively executed fixed prompts at different temperatures. Our results confirm that lower temperatures indeed lead to faster convergence.
- 21 May 2026
- Published by IOS Press
- Vol. 336, 2279-2283
Open AccessConvergence to Steady State in LLM-Generated Ontological Concepts
Large Language Models (LLMs) offer a scalable method to generate candidate concepts for ontology expansion. However, the outputs depend on LLMs’ sampling parameters such as “temperature.” A high temperature leads to increased hallucinations. We consider two hypotheses: (1) LLM outputs for higher temperatures should be supersets of corresponding outputs for lower temperatures; (2) Repeated LLM outputs for lower temperatures should converge to a steady state faster than for higher temperatures. A steady state is defined as k consecutive iterations with no new concepts generated (k-convergence). In this study, we analyzed Environmental Determinants of Health (EnDOH) concepts retrieved at sampling temperatures ranging from 0.1 to 1.0 using concept-structured prompting. Contrary to expectations, higher temperatures do not produce strict supersets of lower temperature outputs. To assess convergence towards a steady state we iteratively executed fixed prompts at different temperatures. Our results confirm that lower temperatures indeed lead to faster convergence.
- Open Access
- 1 January 2026
- Published by Vilnius University Press in Informatica
- p. 229-249
Context-Aware Word-Based Forward Coding
Forward-looking coding has recently been introduced as a source modelling paradigm that exploits predictions of forthcoming symbols. In this paper, we extend this methodology to word-level alphabets, enabling improved compression performance for large and variable-length symbol sets. We present a space-efficient scheme for encoding header information, with particular emphasis on the accurate representation of symbol frequency distributions. In addition, we propose an alternative ordering strategy for word-based dictionaries that leverages the adaptive nature of forward-looking compression. We further show how these techniques can be integrated with a word-based Prediction by Partial Matching model of order one, while avoiding the zero-frequency problem. Experimental results confirm the effectiveness of the proposed approach across multiple datasets.
- 1 January 2026
- Published by Vilnius University Press in Informatica
- p. 229-249
Open AccessContext-Aware Word-Based Forward Coding
Forward-looking coding has recently been introduced as a source modelling paradigm that exploits predictions of forthcoming symbols. In this paper, we extend this methodology to word-level alphabets, enabling improved compression performance for large and variable-length symbol sets. We present a space-efficient scheme for encoding header information, with particular emphasis on the accurate representation of symbol frequency distributions. In addition, we propose an alternative ordering strategy for word-based dictionaries that leverages the adaptive nature of forward-looking compression. We further show how these techniques can be integrated with a word-based Prediction by Partial Matching model of order one, while avoiding the zero-frequency problem. Experimental results confirm the effectiveness of the proposed approach across multiple datasets.
- 18 December 2024
- Published by Springer Nature in Acta Informatica
- Vol. 62 (1) , 1-24
Dense non-binary Fibonacci codes
- 18 December 2024
- Published by Springer Nature in Acta Informatica
- Vol. 62 (1) , 1-24
Dense non-binary Fibonacci codes
- 5 July 2024
- Published by Elsevier in Discrete Applied Mathematics
- Vol. 358, 13-22
Selective dynamic compression
- 5 July 2024
- Published by Elsevier in Discrete Applied Mathematics
- Vol. 358, 13-22
Selective dynamic compression
- 17 March 2023
- Published by Springer Nature in SN Computer Science
- Vol. 4 (3) , 1-12
Weighted Burrows–Wheeler Compression
- 17 March 2023
- Published by Springer Nature in SN Computer Science
- Vol. 4 (3) , 1-12
Weighted Burrows–Wheeler Compression
- 10 January 2023
- Published by Elsevier in Discrete Applied Mathematics
- Vol. 330, 40-50
Bidirectional adaptive compression
- 10 January 2023
- Published by Elsevier in Discrete Applied Mathematics
- Vol. 330, 40-50
Bidirectional adaptive compression
- 9 December 2022
- Published by Elsevier in Discrete Applied Mathematics
- Vol. 327, 18-27
New compression schemes for natural number sequences
- 9 December 2022
- Published by Elsevier in Discrete Applied Mathematics
- Vol. 327, 18-27
New compression schemes for natural number sequences
- Open Access
- 14 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 112043-112052
Generalization of Fibonacci Codes to the Non-Binary Case
Motivated by expected technological developments in which the basic unit of storage might possibly be d -ary elements, with d > 2, and not just bits, we extend the traditional Fibonacci code to non-binary codes of higher order, and prove their theoretical properties, as follows: (1) these codes are fixed in advance , and therefore do not need to be generated for each new probability distribution, yielding very simple and fast encoding and decoding procedures; (2) the codes are prefix-free : no codeword in the code is the prefix of any other codeword; (3) the codes are complete : if one adjoins any other d -ary string as an additional codeword, the obtained extended set of codewords is not uniquely decipherable anymore, and is therefore not useful as it might lead to ambiguities; and (4) the codes provide robustness against decoding errors: the number of lost codewords in case of an error will be limited. An error is defined as a d -ary digit changing its value, or an insertion of an extraneous d -ary digit, or an erroneous deletion of a d -ary digit. We provide experimental results on the compression performance, illustrating that the compression efficiency of non-binary Fibonacci codes is very close to the savings achieved by the corresponding non-binary Huffman coding of the same order, while providing simplicity and robustness.
- 14 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 112043-112052
Open AccessGeneralization of Fibonacci Codes to the Non-Binary Case
Motivated by expected technological developments in which the basic unit of storage might possibly be d -ary elements, with d > 2, and not just bits, we extend the traditional Fibonacci code to non-binary codes of higher order, and prove their theoretical properties, as follows: (1) these codes are fixed in advance , and therefore do not need to be generated for each new probability distribution, yielding very simple and fast encoding and decoding procedures; (2) the codes are prefix-free : no codeword in the code is the prefix of any other codeword; (3) the codes are complete : if one adjoins any other d -ary string as an additional codeword, the obtained extended set of codewords is not uniquely decipherable anymore, and is therefore not useful as it might lead to ambiguities; and (4) the codes provide robustness against decoding errors: the number of lost codewords in case of an error will be limited. An error is defined as a d -ary digit changing its value, or an insertion of an extraneous d -ary digit, or an erroneous deletion of a d -ary digit. We provide experimental results on the compression performance, illustrating that the compression efficiency of non-binary Fibonacci codes is very close to the savings achieved by the corresponding non-binary Huffman coding of the same order, while providing simplicity and robustness.
- 23 April 2022
- Published by Springer Nature in SN Computer Science
- Vol. 3 (3) , 1-10
Combining Forward Compression with PPM
- 23 April 2022
- Published by Springer Nature in SN Computer Science
- Vol. 3 (3) , 1-10
Combining Forward Compression with PPM