Paper Citation Record · LEDGER

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula

As of 16 August 2026, this Paper Citation Record lists 43 of 43 outbound references and 0 inbound Pith citation observations for arXiv:2602.10014.

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

pith.paper-citation-record.v1
2602.10014 v3

Coverage vector

measured 43 of 43 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-03T02:43:38.526645Z

measured 43 of 43 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-16T06:30:59.297886+00:00

measured 0 of 0 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links

measured 0 of 1 external citation measurements

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

Source: cited_works

Reference resolution

43 of 43 outbound references displayed

verified exact0
verified fuzzy0
unresolved43
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 51242c2b-32fb-4315-9ba8-4a9e5f9a5f36 · outbound

This paper cites Intrinsic dimensionality explains the effectiveness of language model fine-tuning.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Intrinsic dimensionality explains the effectiveness of language model fine-tuning

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:33.314701Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1246c825-e4ae-4c10-a5c6-f8e7e9dd842c · outbound

This paper cites Smaller, Weaker, Yet Better: Training LLM Reasoners via Compute-Optimal Sampling.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Smaller, Weaker, Yet Better: Training LLM Reasoners via Compute-Optimal Sampling

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:33.445001Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation baaa6c08-d736-4e3d-ba72-5e8491f35f24 · outbound

This paper cites Training Verifiers to Solve Math Word Problems.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Training Verifiers to Solve Math Word Problems

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:33.551277Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a70328de-abc2-4832-8a73-a421b67040fa · outbound

This paper cites Understanding self-distillation in the presence of label noise.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Understanding self-distillation in the presence of label noise

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:33.684760Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 68b96946-0003-4499-b0d7-c3341c91fb8c · outbound

This paper cites Beyond Model Collapse: Scaling Up with Synthesized Data Requires Verification.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Beyond Model Collapse: Scaling Up with Synthesized Data Requires Verification

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:33.780269Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2cbb0070-04db-4858-aea1-80b079085422 · outbound

This paper cites Self-Consuming Generative Models with Curated Data Provably Optimize Human Preferences.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Self-Consuming Generative Models with Curated Data Provably Optimize Human Preferences

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:33.964748Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 871c7695-7369-4a87-a32b-6d309d392038 · outbound

This paper cites Towards Theoretical Understandings of Self-Consuming Generative Models.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Towards Theoretical Understandings of Self-Consuming Generative Models

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:34.066179Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fea1b8b6-cc38-4cb4-8a42-377033b77326 · outbound

This paper cites Self-verification provably prevents model collapse in recursive synthetic training.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Self-verification provably prevents model collapse in recursive synthetic training

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:34.174752Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ee23ba26-3d4d-45b6-9289-8f71247c0d0f · outbound

This paper cites A Theoretical Perspective: How to Prevent Model Collapse in Self-consuming Training Loops.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula A Theoretical Perspective: How to Prevent Model Collapse in Self-consuming Training Loops

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:34.264756Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 902ae028-2bfa-42e2-88aa-105ebf427232 · outbound

This paper cites Cambridge university press, 2000.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Cambridge university press, 2000

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:34.312806Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7e847d73-d621-438f-83fa-1154dc2ffddb · outbound

This paper cites Is Model Collapse Inevitable? Breaking the Curse of Recursion by Accumulating Real and Synthetic Data.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Is Model Collapse Inevitable? Breaking the Curse of Recursion by Accumulating Real and Synthetic Data

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:34.442154Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bb326200-a19a-47d3-b716-0d536753fb65 · outbound

This paper cites Self-Correcting Self-Consuming Loops for Generative Model Training.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Self-Correcting Self-Consuming Loops for Generative Model Training

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:34.544741Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 69fa0049-0ba1-41dc-8cca-90e6212130fd · outbound

This paper cites The Llama 3 Herd of Models.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula The Llama 3 Herd of Models

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:34.644029Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2a65ce3f-459f-4877-9f61-32c6531bc8d9 · outbound

This paper cites rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinking.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinking

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:34.738415Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ca9bd427-7575-464f-abd9-784748578992 · outbound

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

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:34.983533Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d9e3b4c8-2b19-4b7e-8c88-eaf2aad9f39a · outbound

This paper cites Self-Improvement in Language Models: The Sharpening Mechanism.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Self-Improvement in Language Models: The Sharpening Mechanism

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:35.074111Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a92164a4-cfe1-4d16-8170-e066dbd66c7f · outbound

This paper cites Adastar: Adaptive data sampling for training self-taught reasoners.arXiv preprint arXiv:2505.16322, 2025.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Adastar: Adaptive data sampling for training self-taught reasoners.arXiv preprint arXiv:2505.16322, 2025

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:35.187626Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 999b486e-0963-44e2-9b85-109a71a815a3 · outbound

This paper cites Self-Improving Transformers Overcome Easy-to-Hard and Length Generalization Challenges.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Self-Improving Transformers Overcome Easy-to-Hard and Length Generalization Challenges

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:35.308208Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 17243480-9c70-49c8-a8c3-274f04788f60 · outbound

This paper cites Goedel-Prover: A Frontier Model for Open-Source Automated Theorem Proving.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Goedel-Prover: A Frontier Model for Open-Source Automated Theorem Proving

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:35.409770Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b9d2e9c4-25a8-4196-a655-764c20f43697 · outbound

This paper cites Goedel-Prover-V2: Scaling Formal Theorem Proving with Scaffolded Data Synthesis and Self-Correction.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Goedel-Prover-V2: Scaling Formal Theorem Proving with Scaffolded Data Synthesis and Self-Correction

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:35.496157Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5ab37e65-a95b-4d2c-b9c9-a76742e97a27 · outbound

This paper cites Self-distillation amplifies regular- ization in hilbert space.Advances in Neural Information Processing Systems, 33:3351–3361, 2020.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Self-distillation amplifies regular- ization in hilbert space.Advances in Neural Information Processing Systems, 33:3351–3361, 2020

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:36.035387Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0af07b42-b1e5-46ec-b42b-f12b08630805 · outbound

This paper cites Coherence mechanisms for provable self- improvement.arXiv preprint arXiv:2511.08440, 2025.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Coherence mechanisms for provable self- improvement.arXiv preprint arXiv:2511.08440, 2025

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:36.185231Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 162066db-64c5-4a1a-a34d-48d9c10abfab · outbound

This paper cites Understanding the gains from repeated self-distillation.Advances in Neural Information Processing Systems, 37:7759–7796, 2024.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Understanding the gains from repeated self-distillation.Advances in Neural Information Processing Systems, 37:7759–7796, 2024

Reference 23

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:36.294740Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e8e5813b-eb28-4e05-b35e-49d955d1df80 · outbound

This paper cites DeepSeek-Prover-V2: Advancing Formal Mathematical Reasoning via Reinforcement Learning for Subgoal Decomposition.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula DeepSeek-Prover-V2: Advancing Formal Mathematical Reasoning via Reinforcement Learning for Subgoal Decomposition

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:36.414740Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d3c6978b-91e7-42ca-b492-24a4c81d0505 · outbound

This paper cites Analysing Mathematical Reasoning Abilities of Neural Models.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Analysing Mathematical Reasoning Abilities of Neural Models

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:36.501769Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 33e0d071-2412-44e2-9a11-b8b011ebea37 · outbound

This paper cites A mathematical theory of communication.The Bell system technical journal, 27(3):379–423, 1948.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula A mathematical theory of communication.The Bell system technical journal, 27(3):379–423, 1948

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:36.581709Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5f340510-3d13-47fd-b388-cdd51ecdeb84 · outbound

This paper cites Beyond Human Data: Scaling Self-Training for Problem-Solving with Language Models.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Beyond Human Data: Scaling Self-Training for Problem-Solving with Language Models

Reference 27

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:36.662189Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 88fa9f2e-85a3-466f-8bbc-a2d61ed641f9 · outbound

This paper cites Mind the Gap: Examining the Self-Improvement Capabilities of Large Language Models.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Mind the Gap: Examining the Self-Improvement Capabilities of Large Language Models

Reference 28

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:36.764821Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 02bd9ddc-50c3-4902-8b80-ff068cdc6fa0 · outbound

This paper cites Theoretical modeling of llm self- improvement training dynamics through solver-verifier gap.arXiv preprint arXiv:2507.00075, 2025.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Theoretical modeling of llm self- improvement training dynamics through solver-verifier gap.arXiv preprint arXiv:2507.00075, 2025

Reference 29

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:36.832695Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a8587e93-a18b-4602-b665-802c3ad1ecdb · outbound

This paper cites Canlanguagemodelssolvegraphproblemsinnaturallanguage?Advances in Neural Information Processing Systems, 36:30840–30861, 2023.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Canlanguagemodelssolvegraphproblemsinnaturallanguage?Advances in Neural Information Processing Systems, 36:30840–30861, 2023

Reference 30

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:36.903491Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 676c08f5-ff8e-46d0-b47b-4e2cea3bb481 · outbound

This paper cites Huxley-gödel machine: Human-level coding agent development by an approximation of the optimal self-improving machine, 2025.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Huxley-gödel machine: Human-level coding agent development by an approximation of the optimal self-improving machine, 2025

Reference 31

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:37.006568Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e00b3aec-fe22-4c45-bc02-c247ea87cf25 · outbound

This paper cites Propose, solve, verify: Self-play through formal verification.arXiv preprint arXiv:2512.18160, 2025.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Propose, solve, verify: Self-play through formal verification.arXiv preprint arXiv:2512.18160, 2025

Reference 32

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:37.084920Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0dc1b3f4-a652-4e3b-b1d7-be5c7def22b5 · outbound

This paper cites Probability inequalities for likelihood ratios and convergence rates of sieve mles.The Annals of Statistics, pages 339–362, 1995.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Probability inequalities for likelihood ratios and convergence rates of sieve mles.The Annals of Statistics, pages 339–362, 1995

Reference 33

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:37.203560Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 318dd6b9-3087-411e-8bd5-08c546a795ea · outbound

This paper cites DeepSeek-Prover: Advancing Theorem Proving in LLMs through Large-Scale Synthetic Data.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula DeepSeek-Prover: Advancing Theorem Proving in LLMs through Large-Scale Synthetic Data

Reference 34

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:37.405053Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f95f7968-3759-442c-9096-7548cdb8b70c · outbound

This paper cites DeepSeek-Prover-V1.5: Harnessing Proof Assistant Feedback for Reinforcement Learning and Monte-Carlo Tree Search.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula DeepSeek-Prover-V1.5: Harnessing Proof Assistant Feedback for Reinforcement Learning and Monte-Carlo Tree Search

Reference 35

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:37.574813Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 97c7ebbf-986f-4ad1-9881-97eef9c48816 · outbound

This paper cites Qwen2.5 Technical Report.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Qwen2.5 Technical Report

Reference 36

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:37.684535Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b9071e0e-7881-4e59-8c92-4803d3695144 · outbound

This paper cites Qwen3 Technical Report.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Qwen3 Technical Report

Reference 37

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:37.854813Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5f732155-4010-4193-b5ff-ca448e3493bd · outbound

This paper cites Spendwisely: Maximizing post-training gains in iterative synthetic data bootstrapping.arXiv preprint arXiv:2501.18962, 2025.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Spendwisely: Maximizing post-training gains in iterative synthetic data bootstrapping.arXiv preprint arXiv:2501.18962, 2025

Reference 38

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:37.911309Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c27f7bb9-1692-4506-9202-d39226132c92 · outbound

This paper cites Optimizing Chain-of-Thought Reasoners via Gradient Variance Minimization in Rejection Sampling and RL.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Optimizing Chain-of-Thought Reasoners via Gradient Variance Minimization in Rejection Sampling and RL

Reference 39

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:38.039514Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5d617e96-254c-45fe-97fc-31896e12c58e · outbound

This paper cites Star: Bootstrapping reasoning with reasoning.Advances in Neural Information Processing Systems, 35:15476–15488, 2022.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Star: Bootstrapping reasoning with reasoning.Advances in Neural Information Processing Systems, 35:15476–15488, 2022

Reference 40

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:38.143324Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 95a82f25-5c48-47a9-b428-55d1ef7f0dd4 · outbound

This paper cites B-STaR: Monitoring and Balancing Exploration and Exploitation in Self-Taught Reasoners.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula B-STaR: Monitoring and Balancing Exploration and Exploitation in Self-Taught Reasoners

Reference 41

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:38.300114Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4981f9a3-4db2-4525-89a3-bdfec3acf38e · outbound

This paper cites Leanabell-Prover: Posttraining Scaling in Formal Reasoning.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula Leanabell-Prover: Posttraining Scaling in Formal Reasoning

Reference 42

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:38.423003Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 743d4d8b-b4bd-4634-b99b-3ef94e26d084 · outbound

This paper cites question distribution.

A Task-Centric Theory for Iterative Self-Improvement with Easy-to-Hard Curricula question distribution

Reference 43

Resolution: unresolved; no resolver link, observed 2026-08-03T02:43:38.526645Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

No inbound Pith citation observations are available.