Paper Citation Record · LEDGER

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO

As of 8 August 2026, this Paper Citation Record lists 27 of 27 outbound references and 3 inbound Pith citation observations for arXiv:2505.15694.

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

pith.paper-citation-record.v1
2505.15694 v1

Coverage vector

measured 27 of 27 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-07T15:19:09.839089Z

measured 30 of 30 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-08T06:32:00.761636+00:00

measured 3 of 3 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-07T13:45:05.223275Z

measured 0 of 1 external citation measurements

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

Source: arxiv_reference, observed 2026-05-13T20:13:13.616657Z

Reference resolution

27 of 27 outbound references displayed

verified exact3
verified fuzzy9
unresolved14
parse uncertain0
malformed identifier0
metadata mismatch1

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation cf07d1dc-5adf-4e5b-92bb-781c88a6afce · outbound

This paper cites Claim E.4.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Claim E.4

Reference 1

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:19:11.083198Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 4f432d9e-58b8-45dc-9942-58deb2a88a92 · outbound

This paper cites Thus, by Lemma H.2, we have with probability at least 1 − δ, 1 n nX i=1 ηixi 2 ≤ C · σ · r 1 + ln(1/δ) n , for some universal constant C >0.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Thus, by Lemma H.2, we have with probability at least 1 − δ, 1 n nX i=1 ηixi 2 ≤ C · σ · r 1 + ln(1/δ) n , for some universal constant C >0

Reference 2

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:19:10.976972Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 8ac5c9d9-abc4-42c5-9969-35103f8b9218 · outbound

This paper cites Finally, we use the win rate from these comparisons as our primary performance metric, following the methodology outlined in the DPO paper (Rafailov et al., 2023).

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Finally, we use the win rate from these comparisons as our primary performance metric, following the methodology outlined in the DPO paper (Rafailov et al., 2023)

Reference 5

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:19:11.194972Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 8aa47b56-c5b2-4a2c-ab47-3ca7a5fa914b · outbound

This paper cites Manipulation attacks in local differential privacy.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Manipulation attacks in local differential privacy

Reference 7

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:19:11.868113Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation eb00ceec-6840-4f45-91ae-7ba1ebbef8c7 · outbound

This paper cites Differentially Private Reward Estimation with Preference Feedback.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Differentially Private Reward Estimation with Preference Feedback

Reference 8

Resolution: metadata mismatch; local_arxiv, observed 2026-08-07T15:19:10.315913Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 6f6217b9-2b21-4366-b350-b0d52c2f1c40 · outbound

This paper cites Provably Robust DPO: Aligning Language Models with Noisy Feedback.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Provably Robust DPO: Aligning Language Models with Noisy Feedback

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-07T15:19:08.819245Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 03059725-2090-47c3-94f8-d5d389713b53 · outbound

This paper cites C., Jordan, M.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO C., Jordan, M

Reference 10

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:19:11.734979Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 4e27a73e-4b82-4766-831b-99727d532408 · outbound

This paper cites A tail inequality for quadratic forms of subgaussian random vectors.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO A tail inequality for quadratic forms of subgaussian random vectors

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-07T15:19:09.094011Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f79755dd-1d76-499c-877f-e45cedcb5d0d · outbound

This paper cites Corruption Robust Offline Reinforcement Learning with Human Feedback.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Corruption Robust Offline Reinforcement Learning with Human Feedback

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-07T15:19:09.300383Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d9904036-a23e-4ef2-b25c-ac1113470fd4 · outbound

This paper cites Dueling RL: Reinforcement Learning with Trajectory Preferences.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Dueling RL: Reinforcement Learning with Trajectory Preferences

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-07T15:19:09.343297Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7187fcf9-6b79-4ced-b4d4-3c4c4aa06903 · outbound

This paper cites The importance of online data: Understanding prefer- ence fine-tuning via coverage.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO The importance of online data: Understanding prefer- ence fine-tuning via coverage

Reference 16

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:19:11.617733Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 4c1e7082-13de-40a7-a7a9-ffe05b95f5db · outbound

This paper cites Provable Offline Preference-Based Reinforcement Learning.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Provable Offline Preference-Based Reinforcement Learning

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-07T15:19:09.500623Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e6bcd209-91de-4231-b8dd-d47083cb7456 · outbound

This paper cites Fine-Tuning Language Models from Human Preferences.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Fine-Tuning Language Models from Human Preferences

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-07T15:19:09.536408Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8342ec0c-c248-43a5-81ee-c13432b97416 · outbound

This paper cites Additional Related Work We discuss here more relevant work that do not fit in the main text.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Additional Related Work We discuss here more relevant work that do not fit in the main text

Reference 21

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:19:11.507737Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 7cbe31bb-7cdd-491b-b353-54ac960f42ad · outbound

This paper cites an unresolved cited work.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Unresolved cited work

Reference 22

Resolution: unresolved; raw_fallback, observed 2026-08-07T15:19:11.408244Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation c7fd88bc-2068-45e5-8c2b-5d0552f1d9b5 · outbound

This paper cites rejected.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO rejected

Reference 23

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:19:11.299828Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation a8e615f1-ea4c-4250-8402-d0df3edebdc4 · outbound

This paper cites Chosen” and “Rejected.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Chosen” and “Rejected

Reference 27

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:19:10.866476Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 791c3347-9199-4ea2-90c8-b9eb91a052ad · outbound

This paper cites Robust Reinforcement Learning from Corrupted Human Feedback.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Robust Reinforcement Learning from Corrupted Human Feedback

Reference 1952

Resolution: verified exact; local_arxiv, observed 2026-08-07T15:19:10.582859Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 7b7c3549-e50f-4f13-b26a-6ea1eb994b06 · outbound

This paper cites Exploratory Preference Optimization: Harnessing Implicit Q*-Approximation for Sample-Efficient RLHF.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Exploratory Preference Optimization: Harnessing Implicit Q*-Approximation for Sample-Efficient RLHF

Reference 1965

Resolution: unresolved; no resolver link, observed 2026-08-07T15:19:09.445181Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3e81486b-232c-4b7b-b93b-ba130a1d6d1d · outbound

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

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Correcting the Mythos of KL-Regularization: Direct Alignment without Overoptimization via Chi-Squared Preference Optimization

Reference 2011

Resolution: unresolved; no resolver link, observed 2026-08-07T15:19:09.183433Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e8c9eb8f-a6e9-4b2c-8181-a1e88602ff0c · outbound

This paper cites Exposing Privacy Gaps: Membership Inference Attack on Preference Data for LLM Alignment.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Exposing Privacy Gaps: Membership Inference Attack on Preference Data for LLM Alignment

Reference 2014

Resolution: unresolved; no resolver link, observed 2026-08-07T15:19:09.010064Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f5aa4cf2-f486-4ab7-a49d-81279ea92799 · outbound

This paper cites Classification Under Misspecification: Halfspaces, Generalized Linear Models, and Connections to Evolvability.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Classification Under Misspecification: Halfspaces, Generalized Linear Models, and Connections to Evolvability

Reference 2019

Resolution: verified exact; local_arxiv, observed 2026-08-07T15:19:10.432203Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation b539ee6f-cebb-4b93-93e9-627b6d143fdb · outbound

This paper cites Reinforcement Learning for LLM Post-Training: A Survey.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Reinforcement Learning for LLM Post-Training: A Survey

Reference 2020

Resolution: unresolved; no resolver link, observed 2026-08-07T15:19:09.406012Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 40ae400d-0164-464d-9486-e1460f501434 · outbound

This paper cites Trimmed Maximum Likelihood Estimation for Robust Learning in Generalized Linear Models.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Trimmed Maximum Likelihood Estimation for Robust Learning in Generalized Linear Models

Reference 2021

Resolution: verified exact; local_arxiv, observed 2026-08-07T15:19:10.757471Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 63cf1714-d86f-4816-a113-9bd108e73088 · outbound

This paper cites Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback

Reference 2022

Resolution: unresolved; no resolver link, observed 2026-08-07T15:19:08.018639Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9c8f5672-e659-495c-8784-2e003c483aad · outbound

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

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Value-Incentivized Preference Optimization: A Unified Approach to Online and Offline RLHF

Reference 2023

Resolution: unresolved; no resolver link, observed 2026-08-07T15:19:08.360025Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9bf3d758-8c50-40ea-88e2-3edf4d488e02 · outbound

This paper cites Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback.

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback

Reference 2024

Resolution: unresolved; no resolver link, observed 2026-08-07T15:19:08.255598Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

Observation f1837fc3-a321-403a-b96b-5aad2be5b011 · inbound

Square$\chi$PO: Differentially Private and Robust $\chi^2$-Preference Optimization in Offline Direct Alignment cites this paper.

Square$\chi$PO: Differentially Private and Robust $\chi^2$-Preference Optimization in Offline Direct Alignment A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO

Reference 101

Resolution: unresolved; no resolver link, observed 2026-08-07T13:45:05.223275Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cd1778d0-8b58-41b0-928e-ab3f08ae8001 · inbound

Reinforcement Learning from Human Feedback: A Statistical Perspective cites this paper.

Reinforcement Learning from Human Feedback: A Statistical Perspective A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO

Reference 95

Resolution: verified exact; arxiv_id, observed 2026-05-13T20:13:13.620120Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 0157b7c1-6d1c-441f-87e2-2170d262c79c · inbound

Spurious Correlation Learning in Preference Optimization: Mechanisms, Consequences, and Mitigation via Tie Training cites this paper.

Spurious Correlation Learning in Preference Optimization: Mechanisms, Consequences, and Mitigation via Tie Training A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO

Reference 24

Resolution: metadata mismatch; arxiv_id, observed 2026-05-13T06:32:24.252477Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.