Paper Citation Record · LEDGER

Accelerating RL for LLM Reasoning with Optimal Advantage Regression

As of 11 August 2026, this Paper Citation Record lists 13 of 13 outbound references and 3 inbound Pith citation observations for arXiv:2505.20686.

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

pith.paper-citation-record.v1
2505.20686 v1

Coverage vector

measured 13 of 13 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-07T13:55:57.728078Z

measured 16 of 16 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-10T06:31:04.303077+00:00

measured 3 of 3 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-04T06:47:16.744697Z

measured 0 of 1 external citation measurements

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

Source: arxiv_reference, observed 2026-05-18T08:36:07.284835Z

Reference resolution

13 of 13 outbound references displayed

verified exact0
verified fuzzy7
unresolved6
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation e81a2ab3-e7be-4a5c-b437-cf9bcd404833 · outbound

This paper cites an unresolved cited work.

Accelerating RL for LLM Reasoning with Optimal Advantage Regression Unresolved cited work

Reference 1

Resolution: unresolved; raw_fallback, observed 2026-08-07T13:56:00.295544Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 66aec46c-8ef6-45f2-8d3c-ebff8e96057e · outbound

This paper cites So,a2 = (−2)2 = 4.

Accelerating RL for LLM Reasoning with Optimal Advantage Regression So,a2 = (−2)2 = 4

Reference 2

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:55:59.902129Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation d84a8ba9-54ec-48b2-9408-3179fe2abd2e · outbound

This paper cites an unresolved cited work.

Accelerating RL for LLM Reasoning with Optimal Advantage Regression Unresolved cited work

Reference 3

Resolution: unresolved; raw_fallback, observed 2026-08-07T13:55:58.688420Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 325b6916-991d-49d0-b391-836cad9fdfb5 · outbound

This paper cites So,1 b2 = 1 32 = 1 9.

Accelerating RL for LLM Reasoning with Optimal Advantage Regression So,1 b2 = 1 32 = 1 9

Reference 4

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:56:00.109577Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 5dc3bda0-7723-4f97-9ff4-08425c39f29a · outbound

This paper cites Since3≤b≤5 , the minimum value ofb is 3.

Accelerating RL for LLM Reasoning with Optimal Advantage Regression Since3≤b≤5 , the minimum value ofb is 3

Reference 6

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:55:59.679835Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 4af27d99-0b7b-4316-86b3-b1c4283560c3 · outbound

This paper cites Since−6≤a≤ −2, the minimum value ofa2 is (−2)2 = 4.

Accelerating RL for LLM Reasoning with Optimal Advantage Regression Since−6≤a≤ −2, the minimum value ofa2 is (−2)2 = 4

Reference 7

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:55:59.386030Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 26adfdfb-51e0-40ad-97db-ff2df1d042ab · outbound

This paper cites an unresolved cited work.

Accelerating RL for LLM Reasoning with Optimal Advantage Regression Unresolved cited work

Reference 8

Resolution: unresolved; raw_fallback, observed 2026-08-07T13:55:59.161666Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 42880f05-e596-424c-9dd5-7c7c1c519283 · outbound

This paper cites an unresolved cited work.

Accelerating RL for LLM Reasoning with Optimal Advantage Regression Unresolved cited work

Reference 9

Resolution: unresolved; raw_fallback, observed 2026-08-07T13:55:58.888017Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 8c2c594f-2fde-4bcc-9a59-d248968b7d93 · outbound

This paper cites From these calculations, we see that the greatest possible value is indeed achieved whena=−2andb= 3, giving us: −35 9.

Accelerating RL for LLM Reasoning with Optimal Advantage Regression From these calculations, we see that the greatest possible value is indeed achieved whena=−2andb= 3, giving us: −35 9

Reference 11

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:55:58.506243Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation f814d688-6ba3-4f94-b94b-90668a4e6144 · outbound

This paper cites a+c= 1−1 = 03.

Accelerating RL for LLM Reasoning with Optimal Advantage Regression a+c= 1−1 = 03

Reference 12

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:55:58.263134Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 5ea03061-2779-4809-8854-e3be1289dffc · outbound

This paper cites X y exp(⟨θT+1 , ϕ(x, y)⟩)P y′ exp(⟨θT+1 , ϕ(x, y′)⟩) − exp(⟨θ⋆, ϕ(x, y)⟩)P y′ exp(⟨θ⋆, ϕ(x, y′)⟩) # ≤Ex.

Accelerating RL for LLM Reasoning with Optimal Advantage Regression X y exp(⟨θT+1 , ϕ(x, y)⟩)P y′ exp(⟨θT+1 , ϕ(x, y′)⟩) − exp(⟨θ⋆, ϕ(x, y)⟩)P y′ exp(⟨θ⋆, ϕ(x, y′)⟩) # ≤Ex

Reference 13

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:55:58.018712Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation a25873b9-4bbb-48a7-b9d8-600f23a38b99 · outbound

This paper cites Echo Chamber: RL Post-training Amplifies Behaviors Learned in Pretraining.

Accelerating RL for LLM Reasoning with Optimal Advantage Regression Echo Chamber: RL Post-training Amplifies Behaviors Learned in Pretraining

Reference 2023

Resolution: unresolved; no resolver link, observed 2026-08-07T13:55:56.838631Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7031781b-ba7e-4080-88d5-327e559e5f6e · outbound

This paper cites DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning.

Accelerating RL for LLM Reasoning with Optimal Advantage Regression DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning

Reference 2025

Resolution: unresolved; no resolver link, observed 2026-08-07T13:55:56.722789Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

Observation 6349f898-48a5-4df7-a4d1-6431d7e0e02f · inbound

On the optimization dynamics of RLVR: Gradient gap and step size thresholds cites this paper.

On the optimization dynamics of RLVR: Gradient gap and step size thresholds Accelerating RL for LLM Reasoning with Optimal Advantage Regression

Reference 5

Resolution: verified exact; arxiv_id, observed 2026-05-18T08:36:07.288018Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation b6ff4adb-ef38-43aa-be2c-f7875be47739 · inbound

Variance-Aware Baselines and Adaptive Learning Rates for Reinforcement Learning with Verifiable Rewards cites this paper.

Variance-Aware Baselines and Adaptive Learning Rates for Reinforcement Learning with Verifiable Rewards Accelerating RL for LLM Reasoning with Optimal Advantage Regression

Reference 2025

Resolution: unresolved; no resolver link, observed 2026-08-03T19:38:19.340781Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 62febf29-27b9-4adc-8355-6fec57b6f8b8 · inbound

Variance-Aware Baselines and Adaptive Learning Rates for Reinforcement Learning with Verifiable Rewards cites this paper.

Variance-Aware Baselines and Adaptive Learning Rates for Reinforcement Learning with Verifiable Rewards Accelerating RL for LLM Reasoning with Optimal Advantage Regression

Reference 2025

Resolution: unresolved; no resolver link, observed 2026-08-04T06:47:16.744697Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.