Paper Citation Record · LEDGER

Design Considerations in Offline Preference-based RL

As of 14 August 2026, this Paper Citation Record lists 21 of 21 outbound references and 2 inbound Pith citation observations for arXiv:2502.06861.

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

pith.paper-citation-record.v1
2502.06861 v1

Coverage vector

measured 21 of 21 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-08T19:40:42.080296Z

measured 23 of 23 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-14T06:32:32.682623+00:00

measured 2 of 2 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-05-21T13:06:54.002248Z

measured 0 of 1 external citation measurements

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

Source: arxiv_reference, observed 2026-05-21T13:10:10.404876Z

Reference resolution

21 of 21 outbound references displayed

verified exact0
verified fuzzy1
unresolved20
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 18182556-cba3-4140-bd73-4b789af87ceb · outbound

This paper cites Back to Basics: Revisiting REINFORCE Style Optimization for Learning from Human Feedback in LLMs.

Design Considerations in Offline Preference-based RL Back to Basics: Revisiting REINFORCE Style Optimization for Learning from Human Feedback in LLMs

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:41.977517Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5c30b055-fb90-4bf3-8124-fbb6fabce50c · outbound

This paper cites an unresolved cited work.

Design Considerations in Offline Preference-based RL Unresolved cited work

Reference 3

Resolution: unresolved; raw_fallback, observed 2026-08-08T19:40:42.415603Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.

Observation 89e4aabb-baae-45b5-af0f-9e2d9e9d6917 · outbound

This paper cites Robust Preference Optimization through Reward Model Distillation.

Design Considerations in Offline Preference-based RL Robust Preference Optimization through Reward Model Distillation

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:41.999377Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 28b3bc5c-75f7-4b8e-bf00-8d71f0b107ff · outbound

This paper cites Correcting the Mythos of KL-Regularization: Direct Alignment without Overoptimization via Chi-Squared Preference Optimization.

Design Considerations in Offline Preference-based RL Correcting the Mythos of KL-Regularization: Direct Alignment without Overoptimization via Chi-Squared Preference Optimization

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.004209Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8e476658-38ac-4c5e-8c3c-b49441490b4c · outbound

This paper cites Provably Mitigating Overoptimization in RLHF: Your SFT Loss is Implicitly an Adversarial Regularizer.

Design Considerations in Offline Preference-based RL Provably Mitigating Overoptimization in RLHF: Your SFT Loss is Implicitly an Adversarial Regularizer

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.009598Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9ba1fac7-9632-4932-a5e4-4a15ee9fb486 · outbound

This paper cites Nash Learning from Human Feedback.

Design Considerations in Offline Preference-based RL Nash Learning from Human Feedback

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.019294Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c94c4030-9263-4763-91fc-8508caf05beb · outbound

This paper cites Disentangling Length from Quality in Direct Preference Optimization.

Design Considerations in Offline Preference-based RL Disentangling Length from Quality in Direct Preference Optimization

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.029675Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5f84b809-9c0b-464b-9ae6-7e16dff56c16 · outbound

This paper cites Scaling Laws for Reward Model Overoptimization in Direct Alignment Algorithms.

Design Considerations in Offline Preference-based RL Scaling Laws for Reward Model Overoptimization in Direct Alignment Algorithms

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.034684Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4639f06b-92d8-410e-aac7-ed4c6e12b946 · outbound

This paper cites Gemini: A Family of Highly Capable Multimodal Models.

Design Considerations in Offline Preference-based RL Gemini: A Family of Highly Capable Multimodal Models

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.049982Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a64d54ee-d14d-4af9-a98e-368242dbd36e · outbound

This paper cites Is RLHF More Difficult than Standard RL?.

Design Considerations in Offline Preference-based RL Is RLHF More Difficult than Standard RL?

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.054396Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3c00348f-ff9c-439e-a0ce-a04c3d03c991 · outbound

This paper cites SLiC-HF: Sequence Likelihood Calibration with Human Feedback.

Design Considerations in Offline Preference-based RL SLiC-HF: Sequence Likelihood Calibration with Human Feedback

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.069597Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5641bb44-ced0-44c8-8857-b9c0b6d82cc5 · outbound

This paper cites The optimizer used is Adafactor with learning rate that is constant with a linear warm-up for 2000 steps and a base rate of 1e −.

Design Considerations in Offline Preference-based RL The optimizer used is Adafactor with learning rate that is constant with a linear warm-up for 2000 steps and a base rate of 1e −

Reference 32

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T19:40:42.399596Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.

Observation 02e5e5a0-dffe-4b30-822d-13b599384c68 · outbound

This paper cites Value-Incentivized Preference Optimization: A Unified Approach to Online and Offline RLHF.

Design Considerations in Offline Preference-based RL Value-Incentivized Preference Optimization: A Unified Approach to Online and Offline RLHF

Reference 1952

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:41.989242Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e8739783-b312-4542-905a-f34c8373c5f4 · outbound

This paper cites Calibrating Sequence likelihood Improves Conditional Language Generation.

Design Considerations in Offline Preference-based RL Calibrating Sequence likelihood Improves Conditional Language Generation

Reference 2007

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.064438Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a37fcd4e-f981-4ab9-8bd2-ce2ed2866e01 · outbound

This paper cites On Regularization via Early Stopping for Least Squares Regression.

Design Considerations in Offline Preference-based RL On Regularization via Early Stopping for Least Squares Regression

Reference 2010

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.039660Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation caabea31-3e40-41bf-b61e-2ab71ce38f96 · outbound

This paper cites SimPO: Simple Preference Optimization with a Reference-Free Reward.

Design Considerations in Offline Preference-based RL SimPO: Simple Preference Optimization with a Reference-Free Reward

Reference 2012

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.014455Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d202202d-2a25-428d-938b-5161964e17b7 · outbound

This paper cites KTO: Model Alignment as Prospect Theoretic Optimization.

Design Considerations in Offline Preference-based RL KTO: Model Alignment as Prospect Theoretic Optimization

Reference 2017

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:41.994465Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1daba942-17dd-4a26-bfe4-8731dcd449fd · outbound

This paper cites A Minimaximalist Approach to Reinforcement Learning from Human Feedback.

Design Considerations in Offline Preference-based RL A Minimaximalist Approach to Reinforcement Learning from Human Feedback

Reference 2018

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.044924Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7d1f0652-e606-4e17-8087-744fdbbf5409 · outbound

This paper cites Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation.

Design Considerations in Offline Preference-based RL Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation

Reference 2021

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.059588Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 89d56662-a529-4de4-8ddc-6401e9a9a7d6 · outbound

This paper cites Smaug: Fixing Failure Modes of Preference Optimisation with DPO-Positive.

Design Considerations in Offline Preference-based RL Smaug: Fixing Failure Modes of Preference Optimisation with DPO-Positive

Reference 2022

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:42.024176Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 09c3038c-a61c-4307-9b25-f209c322ef1c · outbound

This paper cites Direct Preference Optimization with an Offset.

Design Considerations in Offline Preference-based RL Direct Preference Optimization with an Offset

Reference 2024

Resolution: unresolved; no resolver link, observed 2026-08-08T19:40:41.983451Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

Observation 3fe1c548-1e61-48a0-9c5f-c0a9d67d8b9e · inbound

Provably avoiding over-optimization in Direct Preference Optimization without knowing the data distribution cites this paper.

Provably avoiding over-optimization in Direct Preference Optimization without knowing the data distribution Design Considerations in Offline Preference-based RL

Reference 1

Resolution: verified exact; arxiv_id, observed 2026-05-16T06:37:28.516891Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.

Observation 6a5f2044-b3f1-4b49-9a79-dc6d9ad3e799 · inbound

Provably avoiding over-optimization in Direct Preference Optimization without knowing the data distribution cites this paper.

Provably avoiding over-optimization in Direct Preference Optimization without knowing the data distribution Design Considerations in Offline Preference-based RL

Reference 1

Resolution: verified exact; arxiv_id, observed 2026-05-21T13:10:10.406878Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-14T06:32:32.682623+00:00.