The Impact of Positional Encoding on Length Generalization in Transformers (2023 · 134 citations)
Transformer language models are usually trained on limited context sizes but are expected to handle longer inputs at inference.
VinePPO: Refining Credit Assignment in RL Training of LLMs (2024 · 63 citations)
When large language models are trained with reinforcement learning on multi-step reasoning, the reward arrives only at the end.

All works

4 works of theirs in our library, the most cited first; 2 with a page here, the others link their original until we summarise them.

The Impact of Positional Encoding on Length Generalization in TransformersConference on Neural Information Processing Systems (NeurIPS)Amirhossein Kazemnejad, Inkit Padhi, Karthikeyan Natesan Ramamurthy, et al.2023 · 134 citations
VinePPO: Refining Credit Assignment in RL Training of LLMsarXiv:2410.01679Amirhossein Kazemnejad, Milad Aghajohari, Eva Portelance, et al.2024 · 63 citations
Paraphrase Generation by Learning How to Edit from Samples ↗Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · no summary yetAmirhossein Kazemnejad, Mohammadreza Salehi, Mahdieh Soleymani Baghshah2020 · 63 citations
DeepSeek-R1 Thoughtology: Let's think about LLM Reasoning ↗arXiv:2504.07128 · no summary yetSara Vera Marjanovic et al.2025 · 47 citations

Co-authors here

6 co-authors with a page, the most papers together first.

Where their works here were published, the most works first, with the editions of those that have a page.

Conference on Neural Information Processing Systems (NeurIPS): NeurIPS 2023
Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics

Papers of our library cite their works 220 times (of the 149,439 papers whose bibliographies we read); 48 citing papers with a page here →.

Updated 2 November 2026