Paper Citation Record · LEDGER

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension

As of 9 August 2026, this Paper Citation Record lists 28 of 28 outbound references and 0 inbound Pith citation observations for arXiv:2608.03494.

A citation records a reference. It does not transfer a finding from one paper to another.

pith.paper-citation-record.v1
2608.03494 v1

Coverage vector

measured 28 of 28 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-05T18:04:09.494559Z

measured 28 of 28 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-09T06:31:02.800959+00:00

measured 0 of 0 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links

measured 0 of 1 external citation measurements

A source-named dated measurement, never combined with another source.

Source: cited_works

Reference resolution

28 of 28 outbound references displayed

verified exact0
verified fuzzy12
unresolved15
parse uncertain0
malformed identifier0
metadata mismatch1

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 74ac2c2a-6cab-4dc2-91a2-ad19ac0cac95 · outbound

This paper cites Neural Machine Translation of Rare Words with Subword Units.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Neural Machine Translation of Rare Words with Subword Units

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-05T18:04:06.609564Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c5ad35bb-6c7b-4f96-92de-37333e19032e · outbound

This paper cites Proceedings of the 2018 conference on empirical methods in natural language processing: System demonstrations , pages=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Proceedings of the 2018 conference on empirical methods in natural language processing: System demonstrations , pages=

Reference 2

Resolution: verified fuzzy; raw_fallback, observed 2026-08-05T18:04:13.491540Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 00a532c5-e026-49df-8f9b-5c1419a9119d · outbound

This paper cites Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing , pages=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing , pages=

Reference 3

Resolution: verified fuzzy; raw_fallback, observed 2026-08-05T18:04:13.245592Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 4f2735ac-1373-446b-9e5b-843271396847 · outbound

This paper cites Advances in neural information processing systems , volume=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Advances in neural information processing systems , volume=

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-05T18:04:07.007445Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f7f6358d-245e-43ea-af95-927d195914e5 · outbound

This paper cites URL: https:/nlp.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension URL: https:/nlp

Reference 5

Resolution: verified fuzzy; raw_fallback, observed 2026-08-05T18:04:12.939084Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 140203a3-0c55-4cd3-8292-50cc87f60bb9 · outbound

This paper cites Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies , pages=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies , pages=

Reference 6

Resolution: verified fuzzy; raw_fallback, observed 2026-08-05T18:04:12.697513Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation eafb42a1-35da-46e4-9c2c-8931218ec1f5 · outbound

This paper cites Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing , pages=.

Reference 7

Resolution: verified fuzzy; raw_fallback, observed 2026-08-05T18:04:12.410049Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation b53b773c-0137-42fb-a546-ad661c6e7beb · outbound

This paper cites International conference on machine learning , pages=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension International conference on machine learning , pages=

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-05T18:04:07.623845Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a0cfbf40-d59c-4354-88d9-45e5fb208895 · outbound

This paper cites MuRIL: Multilingual Representations for Indian Languages.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension MuRIL: Multilingual Representations for Indian Languages

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-05T18:04:07.734213Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6c937ecd-10e0-4e7f-b763-ee88ff987d13 · outbound

This paper cites Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing: Industry Track , pages=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing: Industry Track , pages=

Reference 10

Resolution: verified fuzzy; raw_fallback, observed 2026-08-05T18:04:12.112536Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 9f4df639-bacd-4aa8-afa7-49500daaaaec · outbound

This paper cites Efficient and Effective Text Encoding for Chinese LLaMA and Alpaca.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Efficient and Effective Text Encoding for Chinese LLaMA and Alpaca

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-05T18:04:07.880367Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ef2cb03c-07e0-4b09-a6ca-f8fe6c7ad425 · outbound

This paper cites Adapting Multilingual LLMs to Low-Resource Languages using Continued Pre-training and Synthetic Corpus.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Adapting Multilingual LLMs to Low-Resource Languages using Continued Pre-training and Synthetic Corpus

Reference 12

Resolution: metadata mismatch; local_arxiv, observed 2026-08-05T18:04:09.748618Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 520e40d5-8301-4d9f-b706-480d6a5a5615 · outbound

This paper cites Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , pages=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , pages=

Reference 13

Resolution: verified fuzzy; raw_fallback, observed 2026-08-05T18:04:11.907107Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation a0f2891b-e9be-4dca-b54c-51b749e8f389 · outbound

This paper cites Findings of the Association for Computational Linguistics: EMNLP 2024 , pages=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Findings of the Association for Computational Linguistics: EMNLP 2024 , pages=

Reference 14

Resolution: verified fuzzy; raw_fallback, observed 2026-08-05T18:04:11.629348Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 361a5373-cdf1-411f-8fb7-7365fb25847d · outbound

This paper cites Advances in Neural Information Processing Systems , volume=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Advances in Neural Information Processing Systems , volume=

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-05T18:04:08.262578Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 64d1270a-9562-4f6e-a326-a0eb946718eb · outbound

This paper cites Findings of the Association for Computational Linguistics: NAACL 2024 , pages=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Findings of the Association for Computational Linguistics: NAACL 2024 , pages=

Reference 16

Resolution: verified fuzzy; raw_fallback, observed 2026-08-05T18:04:11.359385Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 9def8cf2-ff97-4c0b-bb8b-a049ec39b31a · outbound

This paper cites AdaptiVocab: Enhancing LLM Efficiency in Focused Domains through Lightweight Vocabulary Adaptation.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension AdaptiVocab: Enhancing LLM Efficiency in Focused Domains through Lightweight Vocabulary Adaptation

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-05T18:04:08.446006Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d274d7bf-7241-40cf-928b-87de2615cee8 · outbound

This paper cites Proceedings of the Fourth Workshop on Multilingual Representation Learning (MRL 2024) , pages=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Proceedings of the Fourth Workshop on Multilingual Representation Learning (MRL 2024) , pages=

Reference 18

Resolution: verified fuzzy; raw_fallback, observed 2026-08-05T18:04:11.026131Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 5fbf3770-9236-4db0-b4c7-eee90f0ac706 · outbound

This paper cites Proceedings of the 2023 conference on empirical methods in natural language processing , pages=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Proceedings of the 2023 conference on empirical methods in natural language processing , pages=

Reference 19

Resolution: verified fuzzy; raw_fallback, observed 2026-08-05T18:04:10.763481Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation e16d83c3-8f87-4b95-a631-bb9bff1f1afb · outbound

This paper cites RedWhale: An Adapted Korean LLM Through Efficient Continual Pretraining.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension RedWhale: An Adapted Korean LLM Through Efficient Continual Pretraining

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-05T18:04:08.718342Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0c95670c-c80e-4f8f-a4d8-e53dd1156534 · outbound

This paper cites Continual Pre-Training for Cross-Lingual LLM Adaptation: Enhancing Japanese Language Capabilities.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Continual Pre-Training for Cross-Lingual LLM Adaptation: Enhancing Japanese Language Capabilities

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-05T18:04:08.811711Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cff7c091-3e1e-4ad8-825c-3b4fa125cd4b · outbound

This paper cites Advances in neural information processing systems , volume=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Advances in neural information processing systems , volume=

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-05T18:04:08.887390Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6e1e6be7-9933-4bc4-b837-ad6ce46f279c · outbound

This paper cites an unresolved cited work.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Unresolved cited work

Reference 23

Resolution: unresolved; raw_fallback, observed 2026-08-05T18:04:10.443566Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 81186828-4a64-4465-8955-7251692e7903 · outbound

This paper cites 2024 , address =.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension 2024 , address =

Reference 24

Resolution: verified fuzzy; raw_fallback, observed 2026-08-05T18:04:10.275986Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 61a24d51-7e72-4256-be47-76456eef77c9 · outbound

This paper cites Gemma 2: Improving Open Language Models at a Practical Size.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Gemma 2: Improving Open Language Models at a Practical Size

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-05T18:04:09.199589Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e6e18b76-1f20-4705-9eb1-1bab47f4f061 · outbound

This paper cites Llama-3-Nanda-10B-Chat: An Open Generative Large Language Model for Hindi.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Llama-3-Nanda-10B-Chat: An Open Generative Large Language Model for Hindi

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-05T18:04:09.305017Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fdb533b5-db61-446c-9d19-874763f0d459 · outbound

This paper cites an unresolved cited work.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Unresolved cited work

Reference 27

Resolution: unresolved; raw_fallback, observed 2026-08-05T18:04:09.985622Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 1f05b4f2-71dc-4bda-ae86-07c8f5c626ee · outbound

This paper cites Transactions of the association for computational linguistics , volume=.

Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension Transactions of the association for computational linguistics , volume=

Reference 28

Resolution: unresolved; no resolver link, observed 2026-08-05T18:04:09.494559Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

No inbound Pith citation observations are available.