Paper Citation Record · LEDGER

On Advantage Estimates for Max@K Policy Gradients

As of 9 August 2026, this Paper Citation Record lists 75 of 75 outbound references and 3 inbound Pith citation observations for arXiv:2606.06080.

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

pith.paper-citation-record.v1
2606.06080 v1

Coverage vector

measured 75 of 75 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-06-28T02:21:57.143016Z

measured 78 of 78 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-09T06:31:02.800959+00:00

measured 3 of 3 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-04T00:49:54.921367Z

measured 0 of 1 external citation measurements

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

Source: cited_works

Reference resolution

75 of 75 outbound references displayed

verified exact41
verified fuzzy0
unresolved30
parse uncertain0
malformed identifier1
metadata mismatch3

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation e9d87534-3b86-49b4-9d36-52120868d367 · outbound

This paper cites Finite-time analysis of the multiarmed bandit problem.Machine learning, 47(2):235–256, 2002.

On Advantage Estimates for Max@K Policy Gradients Finite-time analysis of the multiarmed bandit problem.Machine learning, 47(2):235–256, 2002

Reference 1

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 543eebbd-3c98-410a-ac7e-bfae8c3bea4e · outbound

This paper cites The best of n worlds: Aligning reinforcement learning with best-of-n sampling via max@ k optimisation.

On Advantage Estimates for Max@K Policy Gradients The best of n worlds: Aligning reinforcement learning with best-of-n sampling via max@ k optimisation

Reference 2

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.456387Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 7e4c3140-62f9-4b08-b63f-b3e808a90c04 · outbound

This paper cites Online Preference Alignment for Language Models via Count-based Exploration.

On Advantage Estimates for Max@K Policy Gradients Online Preference Alignment for Language Models via Count-based Exploration

Reference 3

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.459065Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation f0ec27f7-9f27-4d97-b816-ae168c473926 · outbound

This paper cites Post-training as reweighting: A stochastic view of reasoning trajectories in language models.

On Advantage Estimates for Max@K Policy Gradients Post-training as reweighting: A stochastic view of reasoning trajectories in language models

Reference 4

Resolution: metadata mismatch; arxiv_id, observed 2026-07-02T12:16:56.591850Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 2d3a9e82-3707-4987-b5c4-6cf763edcbcd · outbound

This paper cites Exploration by Random Network Distillation.

On Advantage Estimates for Max@K Policy Gradients Exploration by Random Network Distillation

Reference 5

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.450761Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 35de7610-04bb-4d37-98cb-04f19d3a0d84 · outbound

This paper cites arXiv preprint arXiv:2510.15020 , year=.

On Advantage Estimates for Max@K Policy Gradients arXiv preprint arXiv:2510.15020 , year=

Reference 6

Resolution: metadata mismatch; arxiv_id, observed 2026-07-02T12:06:56.385892Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 619fa3af-3bf9-43b9-b1cd-5cbf31564b8a · outbound

This paper cites Evaluating Large Language Models Trained on Code.

On Advantage Estimates for Max@K Policy Gradients Evaluating Large Language Models Trained on Code

Reference 7

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:16:56.588901Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation b02f5bbd-79a9-4cf7-8857-b3f6bccb225e · outbound

This paper cites Pass@k Training for Adaptively Balancing Exploration and Exploitation of Large Reasoning Models.

On Advantage Estimates for Max@K Policy Gradients Pass@k Training for Adaptively Balancing Exploration and Exploitation of Large Reasoning Models

Reference 8

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.383237Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation eb6b85f0-3116-4476-ae8d-a83788160ada · outbound

This paper cites Reasoning with exploration: An entropy perspective.

On Advantage Estimates for Max@K Policy Gradients Reasoning with exploration: An entropy perspective

Reference 9

Resolution: verified exact; doi, observed 2026-06-28T02:31:30.232165Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 02b542aa-6423-49ae-b28c-45706bdeecf5 · outbound

This paper cites Deep reinforcement learning from human preferences.

On Advantage Estimates for Max@K Policy Gradients Deep reinforcement learning from human preferences

Reference 10

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 42035913-4b5e-4044-92b9-ac0aa32f408b · outbound

This paper cites Beyond variance reduction: Understanding the true impact of baselines on policy optimization.

On Advantage Estimates for Max@K Policy Gradients Beyond variance reduction: Understanding the true impact of baselines on policy optimization

Reference 11

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 245a53bb-ddc2-425b-b236-3491c3f5ee93 · outbound

This paper cites Training Verifiers to Solve Math Word Problems.

On Advantage Estimates for Max@K Policy Gradients Training Verifiers to Solve Math Word Problems

Reference 12

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.425930Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 6b214efe-25b4-4ae3-833c-fbbe217175a7 · outbound

This paper cites The Entropy Mechanism of Reinforcement Learning for Reasoning Language Models.

On Advantage Estimates for Max@K Policy Gradients The Entropy Mechanism of Reinforcement Learning for Reasoning Language Models

Reference 13

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:16:56.585783Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 49c68fd0-b753-4d53-8b62-5e7babe84c66 · outbound

This paper cites Weight ensembling improves reasoning in language models.

On Advantage Estimates for Max@K Policy Gradients Weight ensembling improves reasoning in language models

Reference 14

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3c8b4829-1361-4583-beb0-91682e013424 · outbound

This paper cites arXiv preprint arXiv:2505.17621 , year=.

On Advantage Estimates for Max@K Policy Gradients arXiv preprint arXiv:2505.17621 , year=

Reference 15

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.461761Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation fd41b64d-5172-4cca-b828-31ffe3f6ec93 · outbound

This paper cites The Llama 3 Herd of Models.

On Advantage Estimates for Max@K Policy Gradients The Llama 3 Herd of Models

Reference 16

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.443870Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation fdccaf20-1b01-4608-ae6f-ea37d8c6fe86 · outbound

This paper cites Variance reduction techniques for gradient estimates in reinforcement learning.Journal of Machine Learning Research, 5(Nov): 1471–1530, 2004.

On Advantage Estimates for Max@K Policy Gradients Variance reduction techniques for gradient estimates in reinforcement learning.Journal of Machine Learning Research, 5(Nov): 1471–1530, 2004

Reference 17

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ea11b413-dd9c-4a29-b2db-4a27d8b283c9 · outbound

This paper cites MuProp: Unbiased Backpropagation for Stochastic Neural Networks.

On Advantage Estimates for Max@K Policy Gradients MuProp: Unbiased Backpropagation for Stochastic Neural Networks

Reference 18

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.448633Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 30aa64ea-53b9-42ce-bbe1-0fb09374700d · outbound

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

On Advantage Estimates for Max@K Policy Gradients DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning

Reference 19

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.426380Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 83954b2f-cd48-4257-adb2-0de7f8459dc4 · outbound

This paper cites Rewarding the Unlikely: Lifting GRPO Beyond Distribution Sharpening.

On Advantage Estimates for Max@K Policy Gradients Rewarding the Unlikely: Lifting GRPO Beyond Distribution Sharpening

Reference 20

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.428956Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 8a335aa8-db5f-4693-82ea-ad6329f24100 · outbound

This paper cites Measuring Mathematical Problem Solving With the MATH Dataset.

On Advantage Estimates for Max@K Policy Gradients Measuring Mathematical Problem Solving With the MATH Dataset

Reference 21

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.433525Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation c51556c9-30e1-4ed6-bf1e-22f81ba35f0b · outbound

This paper cites A class of statistics with asymptotically normal distribution.

On Advantage Estimates for Max@K Policy Gradients A class of statistics with asymptotically normal distribution

Reference 22

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5aca0c07-0b64-4efe-a780-00e35b22a508 · outbound

This paper cites Emergent Slow Thinking in LLMs as Inverse Tree Freezing.

On Advantage Estimates for Max@K Policy Gradients Emergent Slow Thinking in LLMs as Inverse Tree Freezing

Reference 23

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.448464Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 64be959f-a0ba-4b75-b248-ea764785db81 · outbound

This paper cites arXiv preprint arXiv:2509.25133 , year=.

On Advantage Estimates for Max@K Policy Gradients arXiv preprint arXiv:2509.25133 , year=

Reference 24

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.421477Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 5bab3412-5500-4e49-a263-ee8e2e57402a · outbound

This paper cites Adam: A Method for Stochastic Optimization.

On Advantage Estimates for Max@K Policy Gradients Adam: A Method for Stochastic Optimization

Reference 25

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.431368Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation f8241fdf-c3de-44fd-9bfd-1bdff93c5bd8 · outbound

This paper cites Emergence of exploration in policy gradient reinforcement learning via resetting, 2023.

On Advantage Estimates for Max@K Policy Gradients Emergence of exploration in policy gradient reinforcement learning via resetting, 2023

Reference 26

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f7c6c964-53e6-483f-96f9-7b9a743c4b83 · outbound

This paper cites Hwang, Jiangjiang Yang, Ronan Le Bras, Oyvind Tafjord, Christopher Wilhelm, Luca Soldaini, Noah A.

On Advantage Estimates for Max@K Policy Gradients Hwang, Jiangjiang Yang, Ronan Le Bras, Oyvind Tafjord, Christopher Wilhelm, Luca Soldaini, Noah A

Reference 27

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0a9baef8-00ab-4443-8d09-a9c68b841d2c · outbound

This paper cites Solving quantitative reasoning problems with language models.

On Advantage Estimates for Max@K Policy Gradients Solving quantitative reasoning problems with language models

Reference 28

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 566591fd-f894-43e7-9ac0-cc2a9b78f151 · outbound

This paper cites Jointly Reinforcing Diversity and Quality in Language Model Generations.

On Advantage Estimates for Max@K Policy Gradients Jointly Reinforcing Diversity and Quality in Language Model Generations

Reference 29

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.451211Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation dee092f5-63ba-44e7-b934-48b10a46b4f9 · outbound

This paper cites Can llms guide their own exploration? gradient-guided reinforcement learning for llm reasoning.

On Advantage Estimates for Max@K Policy Gradients Can llms guide their own exploration? gradient-guided reinforcement learning for llm reasoning

Reference 30

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.453670Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 70c5232d-eba7-4788-9178-9635a7407d29 · outbound

This paper cites Understanding r1-zero-like training: A critical perspective.

On Advantage Estimates for Max@K Policy Gradients Understanding r1-zero-like training: A critical perspective

Reference 31

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 740d59f0-ac71-408a-b8e8-e6c15fab080c · outbound

This paper cites RL squeezes, SFT expands: A comparative study of reasoning LLMs.

On Advantage Estimates for Max@K Policy Gradients RL squeezes, SFT expands: A comparative study of reasoning LLMs

Reference 32

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ac8ec50a-cae2-4cbe-b7dd-50064f44a405 · outbound

This paper cites The role of baselines in policy gradient optimization.Advances in Neural Information Processing Systems, 35:17818–17830, 2022.

On Advantage Estimates for Max@K Policy Gradients The role of baselines in policy gradient optimization.Advances in Neural Information Processing Systems, 35:17818–17830, 2022

Reference 33

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fdb5b8d2-32fc-4a8e-a7eb-a6c7ad61f758 · outbound

This paper cites Variational inference for monte carlo objectives.

On Advantage Estimates for Max@K Policy Gradients Variational inference for monte carlo objectives

Reference 34

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a487ccb8-4605-4075-bec3-882270cf3202 · outbound

This paper cites Asynchronous methods for deep reinforce- ment learning.

On Advantage Estimates for Max@K Policy Gradients Asynchronous methods for deep reinforce- ment learning

Reference 35

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c3ff4086-40b9-4c50-8d35-d67f261fd71f · outbound

This paper cites Emergence of exploration in policy gradient reinforcement learning via retrying.

On Advantage Estimates for Max@K Policy Gradients Emergence of exploration in policy gradient reinforcement learning via retrying

Reference 36

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation aaa432fd-26ee-46e7-90a0-daaa390eefc8 · outbound

This paper cites OpenAI o1 System Card.

On Advantage Estimates for Max@K Policy Gradients OpenAI o1 System Card

Reference 37

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.440986Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 068f44fa-383d-40bf-abad-a46d3f9217a8 · outbound

This paper cites Total stochastic gradient algorithms and applications in reinforcement learning.

On Advantage Estimates for Max@K Policy Gradients Total stochastic gradient algorithms and applications in reinforcement learning

Reference 38

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7cc8d450-5098-41bb-ba1a-81b35467b508 · outbound

This paper cites A unified view of likelihood ratio and reparameterization gradients.

On Advantage Estimates for Max@K Policy Gradients A unified view of likelihood ratio and reparameterization gradients

Reference 39

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5f33d517-d2b8-42e5-bed0-3e293bcb355f · outbound

This paper cites PIPPS: Flexible model- based policy search robust to the curse of chaos.

On Advantage Estimates for Max@K Policy Gradients PIPPS: Flexible model- based policy search robust to the curse of chaos

Reference 40

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ccc58204-e901-4294-bf34-6fc3d3bb8c40 · outbound

This paper cites Beyond the Sampled Token: Preserving Candidate Support in RLVR.

On Advantage Estimates for Max@K Policy Gradients Beyond the Sampled Token: Preserving Candidate Support in RLVR

Reference 41

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.431196Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 80b04d8c-869a-437f-896e-e1f698062db6 · outbound

This paper cites Reinforcement learning of motor skills with policy gradients.

On Advantage Estimates for Max@K Policy Gradients Reinforcement learning of motor skills with policy gradients

Reference 42

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation feda5041-2619-462f-bc64-efacbe01732e · outbound

This paper cites Proximal Policy Optimization Algorithms.

On Advantage Estimates for Max@K Policy Gradients Proximal Policy Optimization Algorithms

Reference 43

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.406133Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 978bf51c-3054-476a-bfff-6e3746dd43dd · outbound

This paper cites e3: Learning to Explore Enables Extrapolation of Test-Time Compute for LLMs.

On Advantage Estimates for Max@K Policy Gradients e3: Learning to Explore Enables Extrapolation of Test-Time Compute for LLMs

Reference 44

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.433620Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 283fed06-b34b-4ef7-9aa5-28b5431b5e8f · outbound

This paper cites Rethinking Reflection in Pre-Training.

On Advantage Estimates for Max@K Policy Gradients Rethinking Reflection in Pre-Training

Reference 45

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.418746Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 9c98f150-8c58-49ba-a853-dd94deba054c · outbound

This paper cites DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models.

On Advantage Estimates for Max@K Policy Gradients DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models

Reference 46

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.436184Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation a6ffc73c-cf59-4104-8cc2-e6ef512812c5 · outbound

This paper cites On entropy control in LLM-RL algorithms.

On Advantage Estimates for Max@K Policy Gradients On entropy control in LLM-RL algorithms

Reference 47

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5f9f6c87-a403-485b-b3d9-44f4ae530620 · outbound

This paper cites HybridFlow: A Flexible and Efficient RLHF Framework.

On Advantage Estimates for Max@K Policy Gradients HybridFlow: A Flexible and Efficient RLHF Framework

Reference 48

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.443553Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 04a2bc2c-32e2-40d3-9bb4-51b9022ca6e9 · outbound

This paper cites Outcome-based Exploration for LLM Reasoning.

On Advantage Estimates for Max@K Policy Gradients Outcome-based Exploration for LLM Reasoning

Reference 49

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.413899Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation f3ea7472-d129-4b72-9155-eedbf8a6b1eb · outbound

This paper cites Kakade, Dean Foster, and Udaya Ghai.

On Advantage Estimates for Max@K Policy Gradients Kakade, Dean Foster, and Udaya Ghai

Reference 50

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 56ae5afe-71fd-4cb8-be8d-debda028fceb · outbound

This paper cites Optimizing language models for inference time objectives using reinforcement learning.

On Advantage Estimates for Max@K Policy Gradients Optimizing language models for inference time objectives using reinforcement learning

Reference 51

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c9bdda4e-318e-41e5-a871-8b21caaf1d0a · outbound

This paper cites Rebar: Low-variance, unbiased gradient estimates for discrete latent variable models.Advances in Neural Information Processing Systems, 30, 2017.

On Advantage Estimates for Max@K Policy Gradients Rebar: Low-variance, unbiased gradient estimates for discrete latent variable models.Advances in Neural Information Processing Systems, 30, 2017

Reference 52

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a4c882a5-141e-4c30-8a1d-7f1111eae931 · outbound

This paper cites Representation-Based Exploration for Language Models: From Test-Time to Post-Training.

On Advantage Estimates for Max@K Policy Gradients Representation-Based Exploration for Language Models: From Test-Time to Post-Training

Reference 53

Resolution: verified exact; arxiv_id, observed 2026-07-16T02:22:28.823656Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 04c33fe3-4eb1-414f-b0aa-153b880ba2e8 · outbound

This paper cites Pass@K policy optimization: Solving harder reinforcement learning problems.Advances in Neural Information Processing Systems, 38: 152416–152445, 2025.

On Advantage Estimates for Max@K Policy Gradients Pass@K policy optimization: Solving harder reinforcement learning problems.Advances in Neural Information Processing Systems, 38: 152416–152445, 2025

Reference 54

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 669a75df-30e2-4d7a-831d-0b8a5667faa5 · outbound

This paper cites OctoThinker: Mid-training Incentivizes Reinforcement Learning Scaling.

On Advantage Estimates for Max@K Policy Gradients OctoThinker: Mid-training Incentivizes Reinforcement Learning Scaling

Reference 55

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.418404Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 6f6f9517-3815-4785-b5b3-3c9765a0d875 · outbound

This paper cites The Optimal Reward Baseline for Gradient-Based Reinforcement Learning.

On Advantage Estimates for Max@K Policy Gradients The Optimal Reward Baseline for Gradient-Based Reinforcement Learning

Reference 56

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.445997Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 195d58b3-f851-4598-8de3-4768f6b75741 · outbound

This paper cites Reinforcement Learning with Verifiable Rewards Implicitly Incentivizes Correct Reasoning in Base LLMs.

On Advantage Estimates for Max@K Policy Gradients Reinforcement Learning with Verifiable Rewards Implicitly Incentivizes Correct Reasoning in Base LLMs

Reference 57

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.454152Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 59aaf811-b77f-4c93-81dc-59f67d2d8af0 · outbound

This paper cites Simple statistical gradient-following algorithms for connectionist reinforce- ment learning.Machine Learning, 8(3):229–256, May 1992.

On Advantage Estimates for Max@K Policy Gradients Simple statistical gradient-following algorithms for connectionist reinforce- ment learning.Machine Learning, 8(3):229–256, May 1992

Reference 58

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8792cc16-08d9-48c6-9f54-e21cc4766a3a · outbound

This paper cites Variance Reduction for Policy Gradient with Action-Dependent Factorized Baselines.

On Advantage Estimates for Max@K Policy Gradients Variance Reduction for Policy Gradient with Action-Dependent Factorized Baselines

Reference 59

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.459203Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 2e4e9f0b-bf9f-4747-b751-ac5e31ba3107 · outbound

This paper cites The invisible leash: Why rlvr may or may not escape its origin.

On Advantage Estimates for Max@K Policy Gradients The invisible leash: Why rlvr may or may not escape its origin

Reference 60

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.438710Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation b092c4ec-69cc-4936-a8eb-095eaa57e1d3 · outbound

This paper cites Qwen2.5-Math Technical Report: Toward Mathematical Expert Model via Self-Improvement.

On Advantage Estimates for Max@K Policy Gradients Qwen2.5-Math Technical Report: Toward Mathematical Expert Model via Self-Improvement

Reference 61

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.438752Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 14dd87cb-18e1-40c3-ae63-0b100ed34c89 · outbound

This paper cites Qwen3 Technical Report.

On Advantage Estimates for Max@K Policy Gradients Qwen3 Technical Report

Reference 62

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.456829Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 51bec505-12cd-4214-814a-f74b1d3f61ef · outbound

This paper cites Qwen2.5 Technical Report.

On Advantage Estimates for Max@K Policy Gradients Qwen2.5 Technical Report

Reference 63

Resolution: verified exact; local_arxiv, observed 2026-07-02T12:06:56.446201Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation d725b3bd-2cd2-4efe-beba-0877aaf7be3e · outbound

This paper cites arXiv preprint arXiv:2510.02172 , year=.

On Advantage Estimates for Max@K Policy Gradients arXiv preprint arXiv:2510.02172 , year=

Reference 64

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.403210Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 5f93d10b-abed-4841-b427-b5c443caddea · outbound

This paper cites Does reinforcement learning really incentivize reasoning capacity in LLMs beyond the base model? InThe Thirty-ninth Annual Conference on Neural Information Processing Systems, 2025.

On Advantage Estimates for Max@K Policy Gradients Does reinforcement learning really incentivize reasoning capacity in LLMs beyond the base model? InThe Thirty-ninth Annual Conference on Neural Information Processing Systems, 2025

Reference 65

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 826dfcea-4e34-4dc2-9e96-5a9b3c583ec9 · outbound

This paper cites On the interplay of pre-training, mid-training, and rl on reasoning language models, 2025 a.

On Advantage Estimates for Max@K Policy Gradients On the interplay of pre-training, mid-training, and rl on reasoning language models, 2025 a

Reference 66

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.411254Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation b436eb3d-c72a-4ac2-9093-dd0f0dd13dfb · outbound

This paper cites arXiv preprint arXiv:2509.25810 , year=.

On Advantage Estimates for Max@K Policy Gradients arXiv preprint arXiv:2509.25810 , year=

Reference 67

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.413232Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 066f4021-50c9-41a4-b696-2c794db103c3 · outbound

This paper cites Echo chamber: Rl post-training amplifies behaviors learned in pretraining.

On Advantage Estimates for Max@K Policy Gradients Echo chamber: Rl post-training amplifies behaviors learned in pretraining

Reference 68

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4c242cd4-ec7c-43f0-901d-3edb7e28c52e · outbound

This paper cites Act Only When It Pays: Efficient Reinforcement Learning for LLM Reasoning via Selective Rollouts.

On Advantage Estimates for Max@K Policy Gradients Act Only When It Pays: Efficient Reinforcement Learning for LLM Reasoning via Selective Rollouts

Reference 69

Resolution: metadata mismatch; arxiv_id, observed 2026-07-02T12:06:56.403920Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 024e228f-8c3a-40b3-93a7-ae7df8bbde22 · outbound

This paper cites First Return, Entropy-Eliciting Explore.

On Advantage Estimates for Max@K Policy Gradients First Return, Entropy-Eliciting Explore

Reference 70

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.398549Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation cc922107-1b70-4dd9-9ba4-9d7318dd7fe2 · outbound

This paper cites arXiv preprint arXiv:2509.15194 , year =.

On Advantage Estimates for Max@K Policy Gradients arXiv preprint arXiv:2509.15194 , year =

Reference 71

Resolution: verified exact; arxiv_id, observed 2026-07-02T12:06:56.400294Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 183e932e-84d4-4260-9a22-8f1071cb96e1 · outbound

This paper cites [29] employed a semantic diversity score with an external semantic comparator, and Tuyls et al.

On Advantage Estimates for Max@K Policy Gradients [29] employed a semantic diversity score with an external semantic comparator, and Tuyls et al

Reference 72

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f90062a8-11fd-4ccc-a88a-7c00d637077d · outbound

This paper cites Setlur et al.

On Advantage Estimates for Max@K Policy Gradients Setlur et al

Reference 73

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 855ae7c3-8db9-4d06-9bfe-5b17296af384 · outbound

This paper cites " " Com pu te s bi no mi al c o e f f i c i e n t C (n , k ) in log - space.

On Advantage Estimates for Max@K Policy Gradients " " Com pu te s bi no mi al c o e f f i c i e n t C (n , k ) in log - space

Reference 74

Resolution: unresolved; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8fc7501e-7b95-47e5-986c-f736fdc0740b · outbound

This paper cites BoN mean.

On Advantage Estimates for Max@K Policy Gradients BoN mean

Reference 75

Resolution: malformed identifier; no resolver link, observed 2026-06-28T02:21:57.143016Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

Observation 2ab39a61-7fdc-44b9-ba0c-266f239da232 · inbound

Rank-Conditioned Sample Reuse for the Plackett--Luce Best-of-$K$ Objective cites this paper.

Rank-Conditioned Sample Reuse for the Plackett--Luce Best-of-$K$ Objective On Advantage Estimates for Max@K Policy Gradients

Reference 6

Resolution: unresolved; no resolver link, observed 2026-07-14T06:44:16.198117Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d3a013a1-75d5-4f88-a672-5f78cae46762 · inbound

Multi-Axis Max@K Reinforcement Learning for Representative Diversity in Text-to-Image Generation cites this paper.

Multi-Axis Max@K Reinforcement Learning for Representative Diversity in Text-to-Image Generation On Advantage Estimates for Max@K Policy Gradients

Reference 54

Resolution: unresolved; no resolver link, observed 2026-08-02T00:41:12.582293Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 006768af-557e-43b9-92c8-62ce1a4519a5 · inbound

Stabilized Best-of-$K$ Training for Neural Combinatorial Optimization cites this paper.

Stabilized Best-of-$K$ Training for Neural Combinatorial Optimization On Advantage Estimates for Max@K Policy Gradients

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-04T00:49:54.921367Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.