Typed states for the displayed outbound observations.
Source: paper_references, paper_reference_links, observed 2026-05-19T16:34:47.856606Z
Paper Citation Record · LEDGER
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation
As of 4 August 2026, this Paper Citation Record lists 42 of 42 outbound references and 0 inbound Pith citation observations for arXiv:2605.15239.
A citation records a reference. It does not transfer a finding from one paper to another.
Coverage vector
One-hop event checks from named stored sources.
Source: scholarly_work_events, retraction_status_cache, observed 2026-08-04T06:34:03.388597+00:00
Pith citing papers itemized under the disclosed page cap.
Source: paper_references, paper_reference_links
A source-named dated measurement, never combined with another source.
Source: cited_works
Reference resolution
42 of 42 outbound references displayed
- verified exact2
- verified fuzzy18
- unresolved0
- parse uncertain0
- malformed identifier0
- metadata mismatch22
External citation measurements
No source-named external measurement is stored.
Outbound references
Observation 1647b194-fba2-4184-bf6b-0ec83b0b12aa · outbound
This paper cites Advances in neural information processing systems , volume=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Advances in neural information processing systems , volume=
Reference 1
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.520366Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation b3801cd0-768d-4307-8c7b-0f136a89ba46 · outbound
This paper cites Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback
Reference 2
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.902681Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 35f5cc9c-5bdc-484e-8b07-ee502238aa0f · outbound
This paper cites Jailbreak Attacks and Defenses Against Large Language Models: A Survey.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Jailbreak Attacks and Defenses Against Large Language Models: A Survey
Reference 3
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.882900Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 188ceb5e-da97-47ec-8113-59e95365f06f · outbound
This paper cites Findings of the Association for Computational Linguistics: ACL 2025 , pages=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Findings of the Association for Computational Linguistics: ACL 2025 , pages=
Reference 4
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.518240Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 4dcd14a2-c564-49a4-bcfc-c5eb5f3526c6 · outbound
This paper cites Proceedings of the AAAI Conference on Artificial Intelligence , volume=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Proceedings of the AAAI Conference on Artificial Intelligence , volume=
Reference 5
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.512706Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 624b080a-b7e1-435e-98a5-69f03b29e923 · outbound
This paper cites Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing , pages=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing , pages=
Reference 6
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.514665Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 30fbb816-b491-4503-9819-297fb8ac689a · outbound
This paper cites SafePath: Conformal Prediction for Safe LLM-Based Autonomous Navigation.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation SafePath: Conformal Prediction for Safe LLM-Based Autonomous Navigation
Reference 7
- Resolution
- verified exact
- arxiv_id, observed 2026-05-19T16:37:39.906065Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 7fb23de7-ab3c-457a-863f-1c9f3eab6c0a · outbound
This paper cites Safety-Tuned LLaMAs: Lessons From Improving the Safety of Large Language Models that Follow Instructions.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Safety-Tuned LLaMAs: Lessons From Improving the Safety of Large Language Models that Follow Instructions
Reference 8
- Resolution
- metadata mismatch
- arxiv_id, observed 2026-05-19T16:37:39.913018Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 66d46433-fcf6-4cf6-86b4-75736e1d4b03 · outbound
This paper cites Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!
Reference 9
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.891487Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation e3658083-daf6-45fe-875e-33ddb4234433 · outbound
This paper cites Are Smarter LLMs Safer? Exploring Safety-Reasoning Trade-offs in Prompting and Fine-Tuning.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Are Smarter LLMs Safer? Exploring Safety-Reasoning Trade-offs in Prompting and Fine-Tuning
Reference 10
- Resolution
- metadata mismatch
- arxiv_id, observed 2026-05-19T16:37:39.908935Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation c6a46327-1a16-44ef-bf8b-ca4471668f44 · outbound
This paper cites Safety Tax: Safety Alignment Makes Your Large Reasoning Models Less Reasonable.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Safety Tax: Safety Alignment Makes Your Large Reasoning Models Less Reasonable
Reference 11
- Resolution
- metadata mismatch
- arxiv_id, observed 2026-05-19T16:37:39.880192Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 4bf50f87-b1f6-4e6d-ba63-84ef7768e83e · outbound
This paper cites THINKSAFE: Self-Generated Safety Alignment for Reasoning Models.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation THINKSAFE: Self-Generated Safety Alignment for Reasoning Models
Reference 12
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.899800Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation b8910e1c-87c9-4d1a-ab7a-1ba54351b815 · outbound
This paper cites Safety Alignment Should Be Made More Than Just a Few Tokens Deep.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Safety Alignment Should Be Made More Than Just a Few Tokens Deep
Reference 13
- Resolution
- metadata mismatch
- arxiv_id, observed 2026-05-19T16:37:39.886052Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 645a3552-6d92-4e53-ada0-4ce09a3283af · outbound
This paper cites Self-Distilled Reasoner: On-Policy Self-Distillation for Large Language Models.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Self-Distilled Reasoner: On-Policy Self-Distillation for Large Language Models
Reference 14
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.896866Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 17c18425-8db0-4b32-bb2b-066e5455572e · outbound
This paper cites Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , pages=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , pages=
Reference 15
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.522211Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 0f0c2af1-26f7-4db9-a58d-1711ff60a4a9 · outbound
This paper cites DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
Reference 16
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.888812Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 1219f55d-52ca-4f48-8102-6a6bf30fa058 · outbound
This paper cites Qwen3 Technical Report.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Qwen3 Technical Report
Reference 17
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.894056Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 0e47f25c-cc7b-4631-a008-3c3da78768ff · outbound
This paper cites DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
Reference 18
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.848530Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation bada1515-6f07-4203-8d60-a993c15b7806 · outbound
This paper cites Decoupled Weight Decay Regularization.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Decoupled Weight Decay Regularization
Reference 19
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.877288Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 8c980e12-c1f3-4935-a25d-d519ec05b14e · outbound
This paper cites 2025 , note =.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation 2025 , note =
Reference 20
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.516440Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation c6ac4c7b-8d16-4daa-9e07-761ad7d08e2d · outbound
This paper cites Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations
Reference 21
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.851629Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 54563ba4-806e-4092-bebe-f25238cc18fe · outbound
This paper cites HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 22
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.866341Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 450f1e63-e795-4757-aea0-14fe2f96541c · outbound
This paper cites Advances in Neural Information Processing Systems , volume=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Advances in Neural Information Processing Systems , volume=
Reference 23
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.498831Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation b489457d-4b6b-4fd7-a74a-fd7424122851 · outbound
This paper cites Advances in Neural Information Processing Systems , volume=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Advances in Neural Information Processing Systems , volume=
Reference 24
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.500770Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 489b2817-5b56-4c72-a72c-117c41ce3542 · outbound
This paper cites Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers) , pages=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers) , pages=
Reference 25
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.503054Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 828a5639-82de-4708-bd71-951b7b1521c0 · outbound
This paper cites Advances in neural information processing systems , volume=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Advances in neural information processing systems , volume=
Reference 26
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.504884Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation cf113726-6fe7-40f6-a9b8-c829c7694ea2 · outbound
This paper cites Training Verifiers to Solve Math Word Problems.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Training Verifiers to Solve Math Word Problems
Reference 27
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.863129Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 17eaba70-8051-40eb-b008-0df4b2303c26 · outbound
This paper cites The twelfth international conference on learning representations , year=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation The twelfth international conference on learning representations , year=
Reference 28
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.506689Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation d13beaa8-fe31-4f86-b159-8da330e3692d · outbound
This paper cites GPQA: A Graduate-Level Google-Proof Q&A Benchmark.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation GPQA: A Graduate-Level Google-Proof Q&A Benchmark
Reference 29
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.872110Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 12e4ec2f-4a44-4581-b8cf-6da983dfe552 · outbound
This paper cites Evaluating Large Language Models Trained on Code.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Evaluating Large Language Models Trained on Code
Reference 30
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.845186Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 3654fbbd-a491-4118-bb0b-58bdd1d616fd · outbound
This paper cites Program Synthesis with Large Language Models.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Program Synthesis with Large Language Models
Reference 31
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.854356Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 935d416b-ec31-4a3f-8205-bcbc4af244cc · outbound
This paper cites Bypassing the Safety Training of Open-Source LLMs with Priming Attacks.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Bypassing the Safety Training of Open-Source LLMs with Priming Attacks
Reference 32
- Resolution
- verified exact
- arxiv_id, observed 2026-05-19T16:37:39.857254Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation a3a0dd48-b639-496f-ab77-2e8c2aee7266 · outbound
This paper cites do anything now.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation do anything now
Reference 33
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.492981Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 1ee8f1ee-c151-40b2-a2e9-2eea915cbacc · outbound
This paper cites Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , pages=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , pages=
Reference 34
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.496642Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 0f7ab785-80f3-43d2-9c24-5d2b8724f155 · outbound
This paper cites 2025 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML) , pages=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation 2025 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML) , pages=
Reference 35
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.508591Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 14ad3a0a-248b-40de-82a0-5051fcd8c429 · outbound
This paper cites 2026 , eprint=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation 2026 , eprint=
Reference 36
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.489565Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation f61e7f5e-796d-4f68-92f2-4a33b0e074ba · outbound
This paper cites 2026 , eprint=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation 2026 , eprint=
Reference 37
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.487888Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 505ac6d2-801b-4454-909e-4d28d5cbe0b1 · outbound
This paper cites The twelfth international conference on learning representations , year=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation The twelfth international conference on learning representations , year=
Reference 38
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.491320Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 8b9c96eb-4e73-447f-8edb-bb1d9c7602b1 · outbound
This paper cites The twelfth international conference on learning representations , year=.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation The twelfth international conference on learning representations , year=
Reference 39
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-19T16:37:40.510899Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 856fc587-1272-452a-a13d-cc869e55fd9d · outbound
This paper cites On-Policy Context Distillation for Language Models.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation On-Policy Context Distillation for Language Models
Reference 40
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.874777Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation 2a1119cb-e280-4965-b015-66c67fbee098 · outbound
This paper cites CRISP: Compressed Reasoning via Iterative Self-Policy Distillation.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation CRISP: Compressed Reasoning via Iterative Self-Policy Distillation
Reference 41
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.869050Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Observation eeafd913-7895-4fe5-9c18-a1ca37f899e4 · outbound
This paper cites Universal and Transferable Adversarial Attacks on Aligned Language Models.
Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation Universal and Transferable Adversarial Attacks on Aligned Language Models
Reference 42
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T16:37:39.860061Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-04T06:34:03.388597+00:00.
Pith citing papers
No inbound Pith citation observations are available.