mlsanityData to 7 Oct 2026

Weak-Driven Learning: How Weak Agents make Strong Agents Stronger

arXiv 2602.08222 ↗First posted 9 Feb 2026cs.AIcited by 5

On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks

Abstract (arXiv)

As post-training optimization becomes central to improving large language models, we observe a persistent saturation bottleneck: once models grow highly confident, further training yields diminishing returns. While existing methods continue to reinforce target predictions, we find that informative supervision signals remain latent in models' own historical weak states. Motivated by this observation, we propose WMSS (Weak Agents Can Make Strong Agents Stronger), a post-training paradigm that leverages weak checkpoints to guide continued optimization. By identifying recoverable learning gaps via entropy dynamics and reinforcing them through compensatory learning, WMSS enables strong agents to improve beyond conventional post-training saturation. Experiments on mathematical reasoning and code generation datasets show that agents trained with our approach achieve effective performance improvements, while incurring zero additional inference cost.

Key facts. Cited by 5 papers in the map (data through 7 Oct 2026). Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use. Builds on 41 of its 54 references that are in the map. First posted on arXiv 9 Feb 2026.

Builds on

41 of 54 references.

References · 41

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  3. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  4. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  5. 2021Evaluating Large Language Models Trained on Codecited by 8,053
  6. 2021Program Synthesis with Large Language Modelscited by 3,509
  7. 2024LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Codecited by 1,906
  8. 2020Knowledge Distillation: A Surveycited by 1,578
  9. 2023Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code Generationcited by 1,332
  10. 2023Large Language Models Cannot Self-Correct Reasoning Yetcited by 965
  11. 2021Are NLP Models really able to Solve Simple Math Word Problems?cited by 861
  12. 2017Program Induction by Rationale Generation : Learning to Solve and Explain Algebraic Word Problemscited by 642
29 more
  1. 2024Self-Play Fine-Tuning Converts Weak Language Models to Strong Language Modelscited by 552
  2. 2024BigCodeBench: Benchmarking Code Generation with Diverse Function Calls and Complex Instructionscited by 527
  3. 2019Be Your Own Teacher: Improve the Performance of Convolutional Neural Networks via Self Distillationcited by 491
  4. 2023Reinforced Self-Training (ReST) for Language Modelingcited by 486
  5. 2023Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervisioncited by 445
  6. 2023The False Promise of Imitating Proprietary LLMscited by 254
  7. 2023How Abilities in Large Language Models are Affected by Supervised Fine-tuning Data Compositioncited by 222
  8. 2020ReClor: A Reading Comprehension Dataset Requiring Logical Reasoningcited by 211
  9. 2024On the Self-Verification Limitations of Large Language Models on Reasoning and Planning Taskscited by 137
  10. 2023MiniLLM: On-Policy Distillation of Large Language Modelscited by 129
  11. 2025Self-Evolving Curriculum for LLM Reasoningcited by 107
  12. 2023NEFTune: Noisy Embeddings Improve Instruction Finetuningcited by 99
  13. 2024UNDIAL: Self-Distillation with Adjusted Logits for Robust Unlearning in Large Language Modelscited by 49
  14. 20251.4 Million Open-Source Distilled Reasoning Dataset to Empower Large Language Model Trainingcited by 49
  15. 2024Weak-to-Strong Search: Align Large Language Models via Searching over Small Language Modelscited by 32
  16. 2024Weak-to-Strong Reasoningcited by 32
  17. 2026Your Group-Relative Advantage Is Biasedcited by 32
  18. 2024LLM-Forest: Ensemble Learning of LLMs with Graph-Augmented Prompts for Data Imputationcited by 25
  19. 2024Weak-to-Strong Preference Optimization: Stealing Reward from Weak Aligned Modelcited by 22
  20. 2025Transformer Copilot: Learning from The Mistake Log in LLM Fine-tuningcited by 21
  21. 2024A transfer learning framework for weak-to-strong generalizationcited by 14
  22. 2026Real-Time Aligned Reward Model beyond Semanticscited by 12
  23. 2025GenTool: Enhancing Tool Generalization in Language Models through Zero-to-One and Weak-to-Strong Simulationcited by 9
  24. 2025Semantic Soft Bootstrapping: Long Context Reasoning in LLMs without Reinforcement Learningcited by 9
  25. 2025Incentivizing Strong Reasoning from Weak Supervisioncited by 8
  26. 2025LLMBoost: Make Large Language Models Stronger with Boostingcited by 8
  27. 2024Zero-to-Strong Generalization: Eliciting Strong Capabilities of Large Language Models Iteratively without Gold Labelscited by 4
  28. 2025LightReasoner: Can Small Language Models Teach Large Language Models Reasoning?cited by 3
  29. 2026InstructDiff: Domain-Adaptive Data Selection via Differential Entropy for Efficient LLM Fine-Tuningcited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 5.

5 papers.

  1. 2026Counterfactual Credit Policy Optimization for Multi-Agent Collaborationcited by 10
  2. 2026Adaptive Robust Estimator for Multi-Agent Reinforcement Learningcited by 2
  3. 2026Policy Improvement Reinforcement Learningcited by 2
  4. 2026WDL-OPD: Weak-Driven On-Policy Distillation via Mixture-Constrained Co-Trainingcited by 0
  5. 2026SAKI: Maximal-Coupling-Routed Teacher Supervision for On-Policy Distillationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026When Can LLMs Learn to Reason with Weak Supervision?
  2. 2026Woodpecker Distillation: Weak Models Diagnose Reasoning Bugs in Strong Models
  3. 2026Agentic Systems as Boosting Weak Reasoning Models
  4. 2024Training Agents with Weakly Supervised Feedback from Large Language Models
  5. 2025Alice: Proactive Learning with Teacher's Demonstrations for Weak-to-Strong Generalization
  6. 2026Principled Thoughts for Latent Recursive LLM Systems
  7. 2026Breaking the Capability Ceiling of LLM Post-Training by Reintroducing Markov States
  8. 2025Weak-for-Strong: Training Weak Meta-Agent to Harness Strong Executors