Elior Sulem

4 paper cards in Mova's library · 2018–2025

Elior Sulem wrote 4 of the papers Mova reads, in 2018–2025, most of them on Machine translation; often with Ari Rappoport, Omri Abend, Anita Gelboim.

The profile is the library's match of the name; two people with one name are not always told apart.

Papers

LCHAIM - Investigating Long Context Reasoning in Hebrew (2025)
The authors build LCHAIM, a Hebrew natural language inference set of 8,325 premise–hypothesis pairs, by machine-translating the English ConTRoL dataset.
TafBERTa: Learning Grammatical Rules from Small-Scale Language Acquisition Data in Hebrew (2025)
The paper builds TafBERTa, a small RoBERTa-style masked model (3.3M parameters, 7,317 BPE items) trained from scratch on 233K words of Hebrew child-directed speech from the CHILDES Berman longitudinal corpus, rewritten in Hebrew script and merged with its %mor and %gra annotations (HTBerman: 53K sentences).
Semantic Structural Evaluation for Text Simplification (2018)
The paper asks how to score a simplified sentence when reference simplifications are scarce and not unique.
Simple and Effective Text Simplification Using Semantic and Neural Methods (2018)
The paper adds a semantic splitting step before a neural machine translation simplifier. Direct Semantic Splitting (DSS) parses each sentence with a UCCA semantic parser, which gives scenes (events with their participants), and splits the sentence at its main scenes using two simple rules, one for parallel scenes and…