Pith. sign in

Paper Citation Record · LEDGER

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It

As of 7 August 2026, this Paper Citation Record lists 100 of 148 outbound references and 5 inbound Pith citation observations for arXiv:2505.24119.

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

pith.paper-citation-record.v1
2505.24119 v1

Coverage vector

measured 100 of 148 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-07T12:40:24.573733Z

measured 105 of 105 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-07T06:34:17.273281+00:00

measured 5 of 5 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-06T12:25:21.984374Z

measured 0 of 1 external citation measurements

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

Source: arxiv_reference, observed 2026-07-04T18:10:00.756352Z

Reference resolution

100 of 148 outbound references displayed

  • verified exact7
  • verified fuzzy0
  • unresolved93
  • parse uncertain0
  • malformed identifier0
  • metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 51e1d115-b7a2-4afb-b216-54b5d196d471 · outbound

This paper cites write newline.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It write newline

Reference 1

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:12.806997Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:12.806997Z digest=sha256:2d19f1681385824cdbbde096e3c46b62bd8be14fe49084067d7d1e9a54d74caf

Observation 616a3ad8-c699-44a6-96e6-1e04a6f505f5 · outbound

This paper cites @esa (Ref.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It @esa (Ref

Reference 2

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:12.933128Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:12.933128Z digest=sha256:2642a1617c32f03c11afba91ed4b5acd2948132d1e10e7d26e53b208379e62d0

Observation 737e1d50-a02a-43be-b03d-258a5b77a921 · outbound

This paper cites an unresolved cited work.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Unresolved cited work

Reference 3

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:13.070922Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:13.070922Z digest=sha256:8081e6656437aea1ba62abfe7f6cef2f5e1902a4ba960bf1c7915a0dade730e9

Observation 20b73ef0-a946-4ed6-a1a4-c82352d7f166 · outbound

This paper cites an unresolved cited work.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Unresolved cited work

Reference 4

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:13.220872Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:13.220872Z digest=sha256:b1f0c785bdbdcff1e757d0ebab60c472c70ba64e3b027b069d5a65b95889ccac

Observation eb0d7f95-d1bd-4727-b0e9-9c855cbfcbbc · outbound

This paper cites The multilingual alignment prism: Aligning global and local preferences to reduce harm.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It The multilingual alignment prism: Aligning global and local preferences to reduce harm

Reference 5

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:13.348571Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:13.348571Z digest=sha256:6a9a40b49a6076e6d01e3cc8a476ecc52ecd024cb05abc42c5d9317fd7181ef1

Observation d41d132b-f7af-4542-9f4f-bc2c847ee1f4 · outbound

This paper cites Why knowledge distillation amplifies gender bias and how to mitigate from the perspective of D istil BERT.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Why knowledge distillation amplifies gender bias and how to mitigate from the perspective of D istil BERT

Reference 6

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:13.456374Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:13.456374Z digest=sha256:8511f4949ae1521bc6f3ab34efa0e0d51ac580e1afa1e3231bd3a07145852347

Observation 61d6aa76-1b4f-4a12-998a-cb6e243f91e6 · outbound

This paper cites Jailbreaking LLM s with A rabic transliteration and A rabizi.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Jailbreaking LLM s with A rabic transliteration and A rabizi

Reference 7

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:13.611568Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:13.611568Z digest=sha256:d8ec7998d141ddfd076047223c0f74cda5ad175a60fc892d9f257a92140b43e8

Observation b71c0400-90b9-4c37-ba07-85d19b9dd3e7 · outbound

This paper cites PaLM 2 Technical Report.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It PaLM 2 Technical Report

Reference 8

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:13.761746Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:13.761746Z digest=sha256:b185c3d441b6c1f81ba25e13cf67e08458c62b2569605a869b81e604d92fc1d8

Observation 0d1cbc67-e62a-4ab9-8b73-d7ef4d75dfb6 · outbound

This paper cites Foundational Challenges in Assuring Alignment and Safety of Large Language Models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Foundational Challenges in Assuring Alignment and Safety of Large Language Models

Reference 9

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:13.874731Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:13.874731Z digest=sha256:c5dfef354c537aa34dca49cd018a194a15c60a8dd1c1ccc8bb1adcf04277eedb

Observation 629f2f15-ae3f-4a77-944f-77c0d6d68ce1 · outbound

This paper cites Constitutional AI: Harmlessness from AI Feedback.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Constitutional AI: Harmlessness from AI Feedback

Reference 10

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:14.011410Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:14.011410Z digest=sha256:441caba85be51353b0a104e21022e07a83b1ce244092a2ec42919b1f17d64869

Observation dadf6ce2-3b5f-4f0e-bfeb-244e3d827dc0 · outbound

This paper cites Growing up in a multilingual community: Insights from language socialization.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Growing up in a multilingual community: Insights from language socialization

Reference 11

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:14.134258Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:14.134258Z digest=sha256:e1aa1ba3685891fc5600e60660d14144d8be75e7f5f1af03d83d5579d161d4ad

Observation cbae965c-d0c7-4969-b219-24ad96e225dd · outbound

This paper cites Open Problems in Machine Unlearning for AI Safety.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Open Problems in Machine Unlearning for AI Safety

Reference 12

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:14.232219Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:14.232219Z digest=sha256:51dcf577d063b5e0c25c95fb4929c55062a48000a6dca4dc914b55201a351a49

Observation 8df3cf61-909d-456c-9645-a47db64c28bb · outbound

This paper cites The dark side of language models: Exploring the potential of llms in multimedia disinformation generation and dissemination.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It The dark side of language models: Exploring the potential of llms in multimedia disinformation generation and dissemination

Reference 13

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:14.374461Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:14.374461Z digest=sha256:99c96122e1fe9fcaecefcbe461ca19f31aea9d3f6b1be4c0d5e4a94d5c9c4dcf

Observation 2522f322-ce0e-43d8-9fa8-1dddc2c2799e · outbound

This paper cites The \#benderrule: On naming the languages we study and why it matters.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It The \#benderrule: On naming the languages we study and why it matters

Reference 14

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:14.559818Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:14.559818Z digest=sha256:30519f09a9a77ebdabc9c715485a62e11eb17bdcd71fda0b43c9b58801c815df

Observation 2ca35eaa-84ba-4795-873e-527d224f4617 · outbound

This paper cites On achieving and evaluating language-independence in nlp.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It On achieving and evaluating language-independence in nlp

Reference 15

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:14.726903Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:14.726903Z digest=sha256:a65ea46a43638cd11e1767c5b8f9589a4cba4a46613052a374b431ac4e5f4353

Observation a4b9554c-1306-43f5-be3f-ad2cea117d92 · outbound

This paper cites International ai safety report 2025.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It International ai safety report 2025

Reference 16

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:14.846008Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:14.846008Z digest=sha256:eda7a29cd370038581035c6c236a8381cb651391baefc78c9d32ae3ee764e4a5

Observation 3c4a7bba-4b5f-413b-8aad-f10a190ea32c · outbound

This paper cites S ee GULL multilingual: a dataset of geo-culturally situated stereotypes.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It S ee GULL multilingual: a dataset of geo-culturally situated stereotypes

Reference 17

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:14.989289Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:14.989289Z digest=sha256:fa170a009691eb9814f126f0552b0dcd211df661c86b5b0ef97f3ab94ec400cd

Observation 4c02b775-2776-4ab8-8815-c2f4d5e75f5e · outbound

This paper cites Ai auditing: The broken bus on the road to ai accountability.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Ai auditing: The broken bus on the road to ai accountability

Reference 18

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:15.161815Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:15.161815Z digest=sha256:c1da649fbe9e475ef2f53c04dfbc3fd2aad460dc216973b8dd3ca7acef49e675

Observation df050226-da7e-4a25-815c-751da200b16e · outbound

This paper cites Systematic inequalities in language technology performance across the world`s languages.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Systematic inequalities in language technology performance across the world`s languages

Reference 19

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:15.328653Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:15.328653Z digest=sha256:4d0cfd40e370db53e559665b36febefdcb48dbeda597fe1adb30c27cb281af16

Observation 61418aa8-c539-466b-b0af-e937238df2d6 · outbound

This paper cites Language contamination helps explains the cross-lingual capabilities of E nglish pretrained models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Language contamination helps explains the cross-lingual capabilities of E nglish pretrained models

Reference 20

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:15.443199Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:15.443199Z digest=sha256:b4037649d353927e6123c05e06d7b2de8e465cc2ea65cbd6d6b74308164a91bb

Observation c8a0e7e9-b445-418d-bec7-3a99b10c8d07 · outbound

This paper cites Offensive words in chinese dialects.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Offensive words in chinese dialects

Reference 21

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:15.586061Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:15.586061Z digest=sha256:d09075220471501e91d7c7b02fc45a95c344dbbb64d6f7c4f587c015d732a76e

Observation de4453be-b683-4009-b077-e6136a082176 · outbound

This paper cites Speak Easy: Eliciting Harmful Jailbreaks from LLMs with Simple Interactions.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Speak Easy: Eliciting Harmful Jailbreaks from LLMs with Simple Interactions

Reference 22

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:15.653336Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:15.653336Z digest=sha256:6ab26ce379c72bfd539730b00874d2f91ebb6d8a96328f767e6b0a175fb681f8

Observation 3d80b9c4-ef2f-4cca-87c7-4dfdec963b4a · outbound

This paper cites Gonzalez, Ion Stoica, and Eric P.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Gonzalez, Ion Stoica, and Eric P

Reference 23

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:15.732986Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:15.732986Z digest=sha256:2e1213db8d94773d0cd49d5daa80781c2c0ce10099792fdb5391cae6ab74bd27

Observation b2c98cd4-770f-4725-b0d9-07127430656b · outbound

This paper cites Command A: An Enterprise-Ready Large Language Model.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Command A: An Enterprise-Ready Large Language Model

Reference 24

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:15.783211Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:15.783211Z digest=sha256:ccb67d37b76f0401285074fccc58cc2c87bc97cad1b350227b29630214709ef1

Observation bf3739dc-b0e4-4c80-933b-d5faf1b54620 · outbound

This paper cites Multilingual holistic bias: Extending descriptors and patterns to unveil demographic biases in languages at scale.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Multilingual holistic bias: Extending descriptors and patterns to unveil demographic biases in languages at scale

Reference 25

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:15.871540Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:15.871540Z digest=sha256:7b191f935a86b313989e57386c096a93da97ccf39065c0f91c5037185b1620de

Observation 4e1f2ed6-5cec-49f4-b454-9943f6e04560 · outbound

This paper cites Multilingual holistic bias: Extending descriptors and patterns to unveil demographic biases in languages at scale.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Multilingual holistic bias: Extending descriptors and patterns to unveil demographic biases in languages at scale

Reference 26

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:15.940873Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:15.940873Z digest=sha256:b57e5d447c5fcfd87c984f34485b344780e5af3d7d274464c0034bbb9845ba5e

Observation 8267fe64-7f77-45fc-88b0-71bcbece723e · outbound

This paper cites Toxicity in multilingual machine translation at scale.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Toxicity in multilingual machine translation at scale

Reference 27

Resolution
verified exact
doi, observed 2026-08-07T12:40:30.328424Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

source=arxiv_source observed=2026-08-07T12:40:16.058300Z digest=sha256:8898232d48564ea551a13604640593bf8ad1f23f05cbab73d436e9723bef11cc

Observation d00cd476-491e-4cc2-97ee-74d2b1c41a9d · outbound

This paper cites Overview of the shared task on machine translation gender bias evaluation with multilingual holistic bias.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Overview of the shared task on machine translation gender bias evaluation with multilingual holistic bias

Reference 28

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:16.167678Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:16.167678Z digest=sha256:41869c7276dc0c88eb584f9c44b63a8f369f38fb273514aea907d1fa50146ca5

Observation 45b7565f-9d14-4ea2-a7cf-453fe8a8d013 · outbound

This paper cites Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems

Reference 29

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:16.325876Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:16.325876Z digest=sha256:386321a06ff2894555090dbb4198983c5c22744b5147b4d97f7cc94f62fa85b3

Observation d384d3e5-94b0-4771-a6f4-e6aa0d4160f3 · outbound

This paper cites Aya Vision: Advancing the Frontier of Multilingual Multimodality.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Aya Vision: Advancing the Frontier of Multilingual Multimodality

Reference 30

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:16.433857Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:16.433857Z digest=sha256:4bc5a1fa3938ebfcda5e50e860407e6fab9bf564c531a47fec9f9d7a518f3263

Observation da4604b7-a307-4d68-8f36-2a306bbb5566 · outbound

This paper cites A systematic review of translation and cross-cultural adaptation of instruments for the selection of assistive technologies.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It A systematic review of translation and cross-cultural adaptation of instruments for the selection of assistive technologies

Reference 31

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:16.533841Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:16.533841Z digest=sha256:b6602af1dbe33d3bbae7961b0e290394868b670410d9617eafe37b41f4dec54e

Observation 1c72dfc9-a4ff-4f58-8f74-e4b4b70d49f7 · outbound

This paper cites Emerging Security Challenges of Large Language Models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Emerging Security Challenges of Large Language Models

Reference 32

Resolution
verified exact
local_arxiv, observed 2026-08-07T12:40:32.429547Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

source=arxiv_source observed=2026-08-07T12:40:16.669749Z digest=sha256:2bd052eae81a4a921daec336803ef81e4afac589d06b1b1ed5cb4eef0f68e12e

Observation 3ea1aa87-1a30-473e-92b8-4e5f89d509d8 · outbound

This paper cites Multilingual jailbreak challenges in large language models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Multilingual jailbreak challenges in large language models

Reference 33

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:16.831103Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:16.831103Z digest=sha256:f874318c49e7dc5c83062f6380752842a39e1e31f2901bbd74bbbae496069cef

Observation e7f58f3c-8905-4358-a400-96dc16f4b69a · outbound

This paper cites o rklund, and Henrik Bj \.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It o rklund, and Henrik Bj \

Reference 34

Resolution
verified exact
doi, observed 2026-08-07T12:40:30.199051Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

source=arxiv_source observed=2026-08-07T12:40:16.982060Z digest=sha256:ea87844ed5b42cc284c6aeb14054f2a52a9c6accc76b13d2eadc8499c6cdf303

Observation 169d5098-7717-4221-8c80-4834509368f8 · outbound

This paper cites The emotional force of swearwords and taboo words in the speech of multilinguals.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It The emotional force of swearwords and taboo words in the speech of multilinguals

Reference 35

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:17.099381Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:17.099381Z digest=sha256:5b74ca24198c14496b1a0fd2ac6332fb2438ecabd53fd55aded952bcec999be5

Observation 4f9cddce-7aa6-4b29-9320-08d10e160296 · outbound

This paper cites Attacks, defenses and evaluations for LLM conversation safety: A survey.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Attacks, defenses and evaluations for LLM conversation safety: A survey

Reference 36

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:17.249938Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:17.249938Z digest=sha256:4a60a4ddca4fd6b9b15256511bcacfef5ca916d88d50a1b4a9285985c53371be

Observation f3801be5-8378-428e-8606-20937460b356 · outbound

This paper cites Reducing privacy risks in online self-disclosures with language models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Reducing privacy risks in online self-disclosures with language models

Reference 37

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:17.383368Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:17.383368Z digest=sha256:830ff40bfb4b489eb0e4b1a7aac7a949c58f4cc5fbc40f00952bc34ca8fef439

Observation d052dcc3-1d99-464d-9805-8a80885d1031 · outbound

This paper cites From one to many: Expanding the scope of toxicity mitigation in language models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It From one to many: Expanding the scope of toxicity mitigation in language models

Reference 38

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:17.504203Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:17.504203Z digest=sha256:45eba50c784a47d9c982e051af302b0811db091c8a1ca5141321b5f1cff35269

Observation 6d010448-1c1d-41ae-a2c1-95130241cf2d · outbound

This paper cites Language: The cultural tool.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Language: The cultural tool

Reference 39

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:17.666565Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:17.666565Z digest=sha256:e0b289b410d2ef8cd5c28025cfa4876fb88ada155789d02c3845ea01267a93c8

Observation 17f57367-3c25-4802-bc84-68d70fbc7eae · outbound

This paper cites Sociocultural factors that determine language choice in a multilingual society.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Sociocultural factors that determine language choice in a multilingual society

Reference 40

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:17.796987Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:17.796987Z digest=sha256:20fa707ab3ca49fe67857c2264d109b436de2d8e4ef6db769be933a4b9ad9dd7

Observation f8009d52-807a-41c0-841c-28e8d4dfe84e · outbound

This paper cites Can AI relate: Testing large language model response for mental health support.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Can AI relate: Testing large language model response for mental health support

Reference 41

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:17.907523Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:17.907523Z digest=sha256:7832f4ad383b9f9f0e6248e22b437e025b4814b9cede9debd04614fcecbf312e

Observation 5bd3405e-4f01-4cd4-aea6-d33511186eeb · outbound

This paper cites Code-switching.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Code-switching

Reference 42

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:18.048732Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:18.048732Z digest=sha256:4faf23965f078b5805bd1bc69ce4f0798bd4abc52ed95d9e27edb482d8fc766a

Observation 7e1c755c-734f-4340-9b1c-b8c0a6af8a11 · outbound

This paper cites Regulation and NLP ( R eg NLP ): Taming large language models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Regulation and NLP ( R eg NLP ): Taming large language models

Reference 43

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:18.169319Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:18.169319Z digest=sha256:58aa939f3f3d26914e5fd5282f62a38209059123802b5109e1566edd40c6ce80

Observation 7b29c07f-9333-4b94-89ca-cddc0b996e7b · outbound

This paper cites Language translation & cross-cultural adaptation guideline.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Language translation & cross-cultural adaptation guideline

Reference 44

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:18.290136Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:18.290136Z digest=sha256:cb1dc9adcc8915bed1b3ea20ba14551fd2f079c8f5690486116e07def2a8d93a

Observation 91a852e8-a45c-4c40-aa23-d1fed1c9e40f · outbound

This paper cites The Llama 3 Herd of Models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It The Llama 3 Herd of Models

Reference 45

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:18.411240Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:18.411240Z digest=sha256:33a58f3cde24301f93faee5f8349c705a6313f13d8edc1d173004a17a12cef1a

Observation 9ce7f253-aa27-4ba4-9086-f4463146cef8 · outbound

This paper cites Studying Large Language Model Generalization with Influence Functions.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Studying Large Language Model Generalization with Influence Functions

Reference 46

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:18.546969Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:18.546969Z digest=sha256:fae1a5d7b05e0f32b9519102a4cc52814c1038ac6e3615d4b0e18fbbaac3c389

Observation 87506f66-df2f-47ce-a3e4-7b48fdf1c126 · outbound

This paper cites Deliberative Alignment: Reasoning Enables Safer Language Models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Deliberative Alignment: Reasoning Enables Safer Language Models

Reference 47

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:18.665783Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:18.665783Z digest=sha256:9b0286c84db38dabdf6cb7c17da07532e3b0a94cacab0137649d27e2f5d1bbb4

Observation 487e9b97-4c1a-48e5-9d39-d1ce5961fcd4 · outbound

This paper cites Guerreiro, Duarte M.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Guerreiro, Duarte M

Reference 48

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:18.785294Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:18.785294Z digest=sha256:c58ca84a71cefea228f3bedaf122697614b45bf53b06972bc7436619e83781d7

Observation dd446c42-aea7-4d56-a1e4-1f067c8c076d · outbound

This paper cites Care: Aligning language models for regional cultural awareness.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Care: Aligning language models for regional cultural awareness

Reference 49

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:18.936399Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:18.936399Z digest=sha256:889972f277d349f019ce33de85ea9e49a27f60157bf6975b4a907812a2f98c05

Observation defe0f5e-6c52-449f-a29c-e7f4dbb68bbd · outbound

This paper cites Model merging and safety alignment: One bad model spoils the bunch.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Model merging and safety alignment: One bad model spoils the bunch

Reference 50

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:19.048477Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:19.048477Z digest=sha256:97e0eec7b45d072c8546cb68db893ffbccc5806a36e70e431a862c5666dbad0a

Observation 6285eca5-1934-4db0-8429-068c0a11cd1c · outbound

This paper cites Active learning for robust and representative LLM generation in safety-critical scenarios.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Active learning for robust and representative LLM generation in safety-critical scenarios

Reference 51

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:19.184278Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:19.184278Z digest=sha256:9430cd8a22a89fee47a77bca4dff38414bda9c2aa833089897f1fd2f574b01e9

Observation 7e59be4e-c80a-407f-98bc-38adb3892b44 · outbound

This paper cites TuBA: Cross-Lingual Transferability of Backdoor Attacks in LLMs with Instruction Tuning.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It TuBA: Cross-Lingual Transferability of Backdoor Attacks in LLMs with Instruction Tuning

Reference 52

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:19.323506Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:19.323506Z digest=sha256:5ce489172b23f7ed03c0898a8616c456f0936c82b233af3c285bc87a85df20f2

Observation a1d7c90b-a348-4039-be4b-967631ebd28d · outbound

This paper cites Language in culture; conference on the interrelations of language and other aspects of culture.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Language in culture; conference on the interrelations of language and other aspects of culture

Reference 53

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:19.464070Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:19.464070Z digest=sha256:f2a28d89412dc234c7a9a1dd7942a74000ffaef77e27f3fbae2a34a1cda3e8dd

Observation 008a1799-70e8-42eb-acbd-25d15ba65f96 · outbound

This paper cites T rust A gent: Towards safe and trustworthy LLM -based agents.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It T rust A gent: Towards safe and trustworthy LLM -based agents

Reference 54

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:19.597193Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:19.597193Z digest=sha256:a54ae05ba243f17df47d75bb8ca516acf19304392d2b2cf5a224bb9ee918a7b3

Observation 807e4c5b-4838-4994-83f5-c603f0cbfdfe · outbound

This paper cites A Survey on Large Language Models with Multilingualism: Recent Advances and New Frontiers.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It A Survey on Large Language Models with Multilingualism: Recent Advances and New Frontiers

Reference 55

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:19.717678Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:19.717678Z digest=sha256:feeabc3f16b8e65dec8f2a6d1fb40d464e533dfd3ea37e77dc792c05773ef134

Observation 9235683c-3ab3-4048-907f-a5c3afa2a7ec · outbound

This paper cites MERaLiON-TextLLM: Cross-Lingual Understanding of Large Language Models in Chinese, Indonesian, Malay, and Singlish.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It MERaLiON-TextLLM: Cross-Lingual Understanding of Large Language Models in Chinese, Indonesian, Malay, and Singlish

Reference 56

Resolution
verified exact
local_arxiv, observed 2026-08-07T12:40:32.056607Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

source=arxiv_source observed=2026-08-07T12:40:19.827302Z digest=sha256:1bb8bd9dcb9dd0d804a454cc67ee0b86db67ef7573496fc2b1ec7b1401c1b711

Observation 30e0b36f-7a44-4eb9-af61-b4ba08981cdd · outbound

This paper cites Polyglotoxicityprompts: Multilingual evaluation of neural toxic degeneration in large language models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Polyglotoxicityprompts: Multilingual evaluation of neural toxic degeneration in large language models

Reference 57

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:19.929046Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:19.929046Z digest=sha256:e7cf7fa7ec028ff2abdfdf3f10b5fb2e1bf15818fdd185c8b10c4c56c38a54e4

Observation b05dc424-5d68-43ba-ac7a-a20c44e3dc4b · outbound

This paper cites Safe RLHF-V: Safe Reinforcement Learning from Multi-modal Human Feedback.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Safe RLHF-V: Safe Reinforcement Learning from Multi-modal Human Feedback

Reference 58

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:20.076306Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:20.076306Z digest=sha256:a21875300b1290b5987adf289cb68c2dc1fd6d560b718588750078478e422d84

Observation 36244bda-8f11-4a3c-8f77-6c4fed867af8 · outbound

This paper cites u tze, J \.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It u tze, J \

Reference 59

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:20.235262Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:20.235262Z digest=sha256:5c38e859ca225f396782bd136b1969d81832d9e15e04fb0f5a1c2d37e3e86843

Observation b78cfc23-4555-46ec-9f6a-92ba93fa6f44 · outbound

This paper cites The relationship between culture and language.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It The relationship between culture and language

Reference 60

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:20.377266Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:20.377266Z digest=sha256:d11ad7f19317bc36dcb2a28e04f1c5091193799c8dafaac649a1c9ef7ddeefa2

Observation dadcb9f7-4139-4c41-ab8a-e7860b4be7bb · outbound

This paper cites The state and fate of linguistic diversity and inclusion in the NLP world.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It The state and fate of linguistic diversity and inclusion in the NLP world

Reference 61

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:20.479394Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:20.479394Z digest=sha256:b7f0368ec4d91e42b0bce29b6c0ad38981bd3f26ab03c1c5539bf9eb84d3ff00

Observation 8d695c7c-41cf-46cc-8e41-f9394563ba21 · outbound

This paper cites Towards safe multilingual frontier AI.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Towards safe multilingual frontier AI

Reference 62

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:20.563741Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:20.563741Z digest=sha256:ffdd5bae04f5c362b543278eb61b2ae22626f804f46d3b61daad8802068ba59d

Observation 6fc0cd02-47bf-4a3e-8ca6-85c72853a6b6 · outbound

This paper cites a si, Atte Oksanen, and Pekka R \.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It a si, Atte Oksanen, and Pekka R \

Reference 63

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:20.670738Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:20.670738Z digest=sha256:87ff718bf26172f46d321f6ff82d1ae59295b85d4c804fdc7c93d586fa8df1f9

Observation bd827001-5f22-48f8-8d86-f315492dbdc6 · outbound

This paper cites Banana Bending: Asian-Australian and Asian-Canadian Literatures.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Banana Bending: Asian-Australian and Asian-Canadian Literatures

Reference 64

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:20.771630Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:20.771630Z digest=sha256:7ea2256c63a083adf1d47f33aa26f2fac883ef945126b36fed61299972aa38a1

Observation 79e31604-314d-477a-8493-4496a9af3ccb · outbound

This paper cites Comparative study of multilingual idioms and similes in large language models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Comparative study of multilingual idioms and similes in large language models

Reference 65

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:20.903048Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:20.903048Z digest=sha256:20daef278a79f061db1debaaa43dc931467c1730c28997e3333d00172d230002

Observation 77e3f800-c9fb-4e0e-abf7-3b973bb5fc57 · outbound

This paper cites L ife T ox: Unveiling implicit toxicity in life advice.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It L ife T ox: Unveiling implicit toxicity in life advice

Reference 66

Resolution
verified exact
doi, observed 2026-08-07T12:40:29.961112Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

source=arxiv_source observed=2026-08-07T12:40:21.009859Z digest=sha256:2c58a848c579225b9c9178d0fe5faf9fcd7e43798370464692e54ad0bb62eb61

Observation 7384b7fb-14b5-4bd5-8c07-f3f87201838b · outbound

This paper cites The prism alignment dataset: What participatory, representative and individualised human feedback reveals about the subjective and multicultural alignment of large language models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It The prism alignment dataset: What participatory, representative and individualised human feedback reveals about the subjective and multicultural alignment of large language models

Reference 67

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:21.068074Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:21.068074Z digest=sha256:286a6521834db84d31140898cf58f00dd065ee204d4a852af4fbacf32a6d48aa

Observation 64623f34-035e-45aa-b38e-bd6660f9945e · outbound

This paper cites Untangling hate speech definitions: A semantic componential analysis across cultures and domains.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Untangling hate speech definitions: A semantic componential analysis across cultures and domains

Reference 68

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:21.136750Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:21.136750Z digest=sha256:002f6c7195a965bfed2eb24afc6f449027d03d21865fe9fa0b91766e42c8cf84

Observation 8deed776-fa16-4105-ad57-a8195416292a · outbound

This paper cites Language and culture.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Language and culture

Reference 69

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:21.202410Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:21.202410Z digest=sha256:36934d994cfcbfa937371961959d4e769e432aec5a5d592ae42e6c85464f1b31

Observation f58eed35-e36a-4741-87e9-07eaf5920b54 · outbound

This paper cites D\'ej\`a Vu: Multilingual LLM Evaluation through the Lens of Machine Translation Evaluation.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It D\'ej\`a Vu: Multilingual LLM Evaluation through the Lens of Machine Translation Evaluation

Reference 70

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:21.278210Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:21.278210Z digest=sha256:3265d6c0d0667dcc99893b68079fa9dac31c5451a5b4371f52cb8eb0e51622f8

Observation 0276d684-4b4f-4178-8686-d83e59baa78a · outbound

This paper cites LLM -based synthetic datasets: Applications and limitations in toxicity detection.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It LLM -based synthetic datasets: Applications and limitations in toxicity detection

Reference 71

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:21.377864Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:21.377864Z digest=sha256:5e31cdd8331a4758500d84afe085afca3e373c1928246f4cc076962a70e61106

Observation 211d32bb-4dbe-4dda-a7c1-c4f77be5fff0 · outbound

This paper cites Specific versus General Principles for Constitutional AI.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Specific versus General Principles for Constitutional AI

Reference 72

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:21.447285Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:21.447285Z digest=sha256:13e4d0b310273c0558f3ce034e725f7c5e2151abc28ed5f96373b73302b666ef

Observation 41e70d9d-ad99-4615-8b52-f7b86d65158a · outbound

This paper cites C3ai: Crafting and evaluating constitutions for constitutional ai.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It C3ai: Crafting and evaluating constitutions for constitutional ai

Reference 73

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:21.543409Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:21.543409Z digest=sha256:4969b46cb62d3eaefb32f59bc3b3a9fceb776fc56d9495056bd9eb35d0a2e29b

Observation f31fff6f-2dda-4f41-8736-86caf7e135b6 · outbound

This paper cites Ai value alignment: How we can align artificial intelligence with human values.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Ai value alignment: How we can align artificial intelligence with human values

Reference 74

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:21.642305Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:21.642305Z digest=sha256:8973b865a3d172be19e04c660826cb63e4894ba2605e9b5411844c6e947b8d88

Observation 212698ae-2077-42f7-bd6c-2951ad69ffc9 · outbound

This paper cites Preference tuning for toxicity mitigation generalizes across languages.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Preference tuning for toxicity mitigation generalizes across languages

Reference 75

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:21.763988Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:21.763988Z digest=sha256:62db12a2f4c12ebb300798f6b27d952e044840034ae15f99d83d66c545858e74

Observation 7745b8cb-beb6-4f46-9ae1-25b4f624afce · outbound

This paper cites Beyond Single-Turn: A Survey on Multi-Turn Interactions with Large Language Models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Beyond Single-Turn: A Survey on Multi-Turn Interactions with Large Language Models

Reference 76

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:21.915386Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:21.915386Z digest=sha256:1dee9cf20ff9d0dfb264b94c6e777a70775eebcaffe807abf69c104b7b61be5d

Observation 28df8f69-88a0-4eaf-9c6f-bb07ba0b6969 · outbound

This paper cites MaLA-500: Massive Language Adaptation of Large Language Models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It MaLA-500: Massive Language Adaptation of Large Language Models

Reference 77

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:22.060872Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:22.060872Z digest=sha256:0f08bba097e20a24416c195b578745c427903d11a7c13eb296a908c4166a51ba

Observation 86af2a50-ed85-41be-86b9-2390d8c629bb · outbound

This paper cites Systematicity, compositionality and transitivity of deep NLP models: a metamorphic testing perspective.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Systematicity, compositionality and transitivity of deep NLP models: a metamorphic testing perspective

Reference 78

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:22.140560Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:22.140560Z digest=sha256:84cd438747ed7fcc6b5aee70ffe0e53b35fb692926cdb455721429963d9f31a0

Observation 36d8c2a0-77ca-47ec-bbdf-bc1a3c030dd4 · outbound

This paper cites Translation and cultural adaptation of the survey instruments.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Translation and cultural adaptation of the survey instruments

Reference 79

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:22.199426Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:22.199426Z digest=sha256:dac7c3a6baa6a764e203d7ff05279ca2da8aed884c63fb6fb4aad97895897d39

Observation a7e9aebd-bf32-49b5-9a34-1331f38eade4 · outbound

This paper cites Language and culture.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Language and culture

Reference 80

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:22.245761Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:22.245761Z digest=sha256:ec834df0d6a42d9ad3c556f7e2e2c71c31f825ae2b7a5c580b86d72c1d121ceb

Observation c295a380-18f9-457d-8eb0-2f17484320c3 · outbound

This paper cites Translation, cross-cultural adaptation and psychometric evaluation of yoruba version of the short-form 36 health survey.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Translation, cross-cultural adaptation and psychometric evaluation of yoruba version of the short-form 36 health survey

Reference 81

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:22.297819Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:22.297819Z digest=sha256:1f9365c007cff0f508d3f48c792d31592cd8cc03feadb0198e64593305211969

Observation 988e6fce-7a85-466c-93a0-11c97f078dd9 · outbound

This paper cites COVID -19 in B ulgarian social media: Factuality, harmfulness, propaganda, and framing.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It COVID -19 in B ulgarian social media: Factuality, harmfulness, propaganda, and framing

Reference 82

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:22.360739Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:22.360739Z digest=sha256:9ba9daf435233e7236cc3dd33683de899541d77caafc5a58611a3b60d52bce86

Observation 44666680-0793-4742-be06-75192e1d60d1 · outbound

This paper cites Large language models in healthcare and medical domain: A review.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Large language models in healthcare and medical domain: A review

Reference 83

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:22.447277Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:22.447277Z digest=sha256:940db073bd255cbfc09401d1169264ed3a5ce5b9e5f5fdf2201e0d833793eeb6

Observation b4e5f02c-2af4-4f5d-b59a-7582dba16130 · outbound

This paper cites An llm-driven chatbot in higher education for databases and information systems.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It An llm-driven chatbot in higher education for databases and information systems

Reference 84

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:22.511905Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:22.511905Z digest=sha256:7f9043bbb6400c0159eefd3637bf6e21a9c4de163bbf5b1e9af8a310f1312340

Observation 99c81787-e02b-4529-834e-4b8ca51e0368 · outbound

This paper cites “i searched for a religious song in amharic and got sexual content instead’’: Investigating online harm in low-resourced languages on youtube.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It “i searched for a religious song in amharic and got sexual content instead’’: Investigating online harm in low-resourced languages on youtube

Reference 85

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:22.619595Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:22.619595Z digest=sha256:84202e9fee0062a089f3e1683f67adc79bd69f7577ed9fd6b160441d2a81f8f0

Observation d744b5b1-ff42-45d1-8291-488083961c68 · outbound

This paper cites The zeno`s paradox of low-resource' languages.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It The zeno`s paradox of low-resource' languages

Reference 86

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:22.735090Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:22.735090Z digest=sha256:a115f08577ff6d1e1ebd0164a4aa0210f0acbf71cc4572966f71e7ecf5a76dda

Observation 8bd3420d-b2e8-484e-991c-3cb9c9750e1c · outbound

This paper cites code switching.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It code switching

Reference 87

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:22.855992Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:22.855992Z digest=sha256:114d35fbe5d074faaf918a0559cd15459f556fcf6b8cb29292cc4fa285bf5d81

Observation a707d1d5-2aec-4ef5-beb2-4dd3158f6511 · outbound

This paper cites Niraula, Saurab Dulal, and Diwa Koirala.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Niraula, Saurab Dulal, and Diwa Koirala

Reference 88

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:23.005650Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:23.005650Z digest=sha256:62912b064ee8e03244bd7254b527b87230caffabf8ad6c5634d85e101a04fdda

Observation b5718765-62d7-4330-b367-dc4a33c582c8 · outbound

This paper cites Exposing the limits of zero-shot cross-lingual hate speech detection.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Exposing the limits of zero-shot cross-lingual hate speech detection

Reference 89

Resolution
verified exact
doi, observed 2026-08-07T12:40:29.732463Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

source=arxiv_source observed=2026-08-07T12:40:23.141802Z digest=sha256:81aa00a1716cc4de3d029482be6afebbc5d10f4aff8f4b555d0bbe71f90e0cd4

Observation 3fbd7183-6e72-49f6-b2b5-bd49d6434833 · outbound

This paper cites Towards ai accountability infrastructure: Gaps and opportunities in ai audit tooling.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Towards ai accountability infrastructure: Gaps and opportunities in ai audit tooling

Reference 90

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:23.232135Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:23.232135Z digest=sha256:283c4e23a26840f3b4671f045abeba3dd5cb70b050a1d2bb46e2706fcef1dda0

Observation 3236720b-8dbd-4cf8-b32e-2139af75ea63 · outbound

This paper cites Introducing ChatGPT.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Introducing ChatGPT

Reference 91

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:23.314400Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:23.314400Z digest=sha256:b8c60fa0d0c5b0820e8ec869619edad25f09459f748d9f53ba718e0c10efed98

Observation e79596c7-4ac4-4e76-a918-f08c9296125b · outbound

This paper cites Openai gpt-4.5 system card.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Openai gpt-4.5 system card

Reference 92

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:23.409123Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:23.409123Z digest=sha256:74a07c46ab0eb716a3ab77efefc9d699b15a1d42b93c80a94bd4d3c98e74ccd2

Observation 03debe68-8f4e-41fb-93a0-08d021473f9d · outbound

This paper cites Gemini goes to M ed school: Exploring the capabilities of multimodal large language models on medical challenge problems & hallucinations.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Gemini goes to M ed school: Exploring the capabilities of multimodal large language models on medical challenge problems & hallucinations

Reference 93

Resolution
verified exact
doi, observed 2026-08-07T12:40:29.566379Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

source=arxiv_source observed=2026-08-07T12:40:23.495470Z digest=sha256:a7c3c7c5f9462ec09c8bfa87bd0a50a0439c19fcf3ffe8dcf840d31112bf3504

Observation 34eeb2f2-0709-4cf8-b501-52eacbcccde1 · outbound

This paper cites Towards multidomain and multilingual abusive language detection: a survey.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Towards multidomain and multilingual abusive language detection: a survey

Reference 94

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:23.654465Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:23.654465Z digest=sha256:0c8520736fecf3bf4a6f8a322fab2381e4f8c2b7823c814ca0727684c8479ac3

Observation 4d648ca8-95f6-4ecc-993d-01abd12d5854 · outbound

This paper cites The Multilingual Divide and Its Impact on Global AI Safety.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It The Multilingual Divide and Its Impact on Global AI Safety

Reference 95

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:23.775161Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:23.775161Z digest=sha256:fb125c72751b062a19ce12c891725b51f73b7e706d22c57058ee36e0ce66a89f

Observation e439ac4e-8417-430b-b201-d9bf4e5d4f10 · outbound

This paper cites The brazilian version of start back screening tool-translation, cross-cultural adaptation and reliability.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It The brazilian version of start back screening tool-translation, cross-cultural adaptation and reliability

Reference 96

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:23.910144Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:23.910144Z digest=sha256:46e6e81e9e6cfa72e0d4116a9dd10d02dba77b38b72ccec269f3a73e4e29f250

Observation 8efdd8de-1ba6-4bcf-bd6c-7b11280e8ecd · outbound

This paper cites Civics: Building a dataset for examining culturally-informed values in large language models.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Civics: Building a dataset for examining culturally-informed values in large language models

Reference 97

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:24.049791Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:24.049791Z digest=sha256:9651130e27753f14e5d230bd928d2ec0adb33d910bc222e795d1d72c54d5d5c4

Observation 7b27157b-b202-4d8f-a9dd-08b75d0636cb · outbound

This paper cites Countering hateful and offensive speech online - open challenges.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Countering hateful and offensive speech online - open challenges

Reference 98

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:24.208273Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:24.208273Z digest=sha256:4810687330588f35d52e9fd3a1345f6e1abf437de783ad62cb23fc7ffda8c499

Observation f705194e-bdc7-4622-908f-517264c58dfa · outbound

This paper cites Towards understanding the fragility of multilingual llms against fine-tuning attacks.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Towards understanding the fragility of multilingual llms against fine-tuning attacks

Reference 99

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:24.416262Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:24.416262Z digest=sha256:2f0a39ed145b325919b435906f81e251ef46572685795e5043bccc2c693456e0

Observation 6b9ac04d-f6fa-498d-b142-8cebee39b1c6 · outbound

This paper cites Multilingual Large Language Model: A Survey of Resources, Taxonomy and Frontiers.

The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It Multilingual Large Language Model: A Survey of Resources, Taxonomy and Frontiers

Reference 100

Resolution
unresolved
no resolver link, observed 2026-08-07T12:40:24.573733Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-07T12:40:24.573733Z digest=sha256:e61d8b80c9c0bd347e1498f416864a30f0a7349ba9cf01cb6864e2a8000ce34c

Pith citing papers

Observation b0c2cb60-478d-434b-8fe9-97faa93358d3 · inbound

The Problem with Safety Classification is not just the Models cites this paper.

The Problem with Safety Classification is not just the Models The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It

Reference 14

Resolution
unresolved
no resolver link, observed 2026-08-06T12:25:21.984374Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=pdf_text observed=2026-08-06T12:25:21.984374Z digest=sha256:6b96701de5b2d352167694129b0d93893c19ba6ae0612846ade883a7a7e37b5c

Observation 9052e378-89fe-4e1f-b251-4b67082e5254 · inbound

Response-Based Knowledge Distillation for Multilingual Jailbreak Prevention Unwittingly Compromises Safety cites this paper.

Response-Based Knowledge Distillation for Multilingual Jailbreak Prevention Unwittingly Compromises Safety The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It

Reference 50

Resolution
verified exact
arxiv_id, observed 2026-05-17T01:28:48.794515Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

source=pdf_text observed=2026-05-17T01:27:16.967080Z digest=sha256:6a0031d1fb4c691365e26e6415a4e3ef061be6d1f3f7cb9bf23eb5450edae481

Observation a107d0f0-79de-4aff-b2be-6745c817da34 · inbound

ROK-FORTRESS: Measuring the Effect of Geopolitical Transcreation for National Security and Public Safety cites this paper.

ROK-FORTRESS: Measuring the Effect of Geopolitical Transcreation for National Security and Public Safety The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It

Reference 23

Resolution
unresolved
no resolver link, observed 2026-07-12T16:49:32.395383Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=pdf_text observed=2026-07-12T16:49:32.395383Z digest=sha256:bc89535ef67a9eef7afc4dcd22bb3b8390d42f6c431fec1cca8dd55198ea74a9

Observation bcdfc851-0ae3-4bdf-bef9-8d0e6d9204cf · inbound

SomaliBench Eval: Measuring English-to-Somali Refusal Gaps in Open-Weight Language Models cites this paper.

SomaliBench Eval: Measuring English-to-Somali Refusal Gaps in Open-Weight Language Models The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It

Reference 13

Resolution
verified exact
arxiv_id, observed 2026-06-29T22:34:01.317730Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

source=pdf_text observed=2026-06-29T22:33:37.250679Z digest=sha256:2b360bde39b345406bec6adf6ba35b5acc593af6ee9270214f58df7fee89dd91

Observation e9443d60-1db5-4c22-a45a-6ba17ad0474c · inbound

Multilingual Refusal Alignment for Safer Large Language Models cites this paper.

Multilingual Refusal Alignment for Safer Large Language Models The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating It

Reference 70

Resolution
verified exact
arxiv_id, observed 2026-07-04T18:10:00.758266Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

source=arxiv_source observed=2026-07-04T18:09:17.531727Z digest=sha256:af2c610a02190fa1fee1c56b7897c22326b71632c453b20fa6bf36b164436b2a