mlsanityData to 14 Oct 2026

Beyond path selection: Better LLMs for Scientific Information Extraction with MimicSFT and Relevance and Rule-induced(R2)GRPO

arXiv 2505.22068 ↗First posted 28 May 2025cs.CL cs.AI cs.IRcited by 2

On the map: LLM reasoning → Knowledge- and tool-augmented reasoning

Abstract (arXiv)

Previous study suggest that powerful Large Language Models (LLMs) trained with Reinforcement Learning with Verifiable Rewards (RLVR) only refines reasoning path without improving the reasoning capacity in math tasks while supervised-finetuning(SFT) with distillation can. We study this from the view of Scientific information extraction (SciIE) where LLMs and reasoning LLMs underperforms small Bert-based models. SciIE require both the reasoning and memorization. We argue that both SFT and RLVR can refine the reasoning path and improve reasoning capacity in a simple way based on SciIE. We propose two-stage training with 1. MimicSFT, using structured reasoning templates without needing high-quality chain-of-thought data, 2. R2GRPO with relevance and rule-induced rewards. Experiments on scientific IE benchmarks show that both methods can improve the reasoning capacity. R2GRPO with mimicSFT surpasses baseline LLMs and specialized supervised models in relation extraction. Our code is available at https://github.com/ranlislz/R2GRPO.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in Knowledge- and tool-augmented reasoning, one of 17 directions in LLM reasoning. Builds on 30 of its 46 references that are in the map. First posted on arXiv 28 May 2025.

Builds on

30 of 46 references.

References · 30

  1. 2023GPT-4 Technical Reportcited by 18,396
  2. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  3. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  4. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  5. 2020Scaling Laws for Neural Language Modelscited by 6,586
  6. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,330
  7. 2024Qwen2.5 Technical Reportcited by 4,942
  8. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,780
  9. 2022Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedbackcited by 3,720
  10. 2022OPT: Open Pre-trained Transformer Language Modelscited by 3,512
  11. 2024Gemini 1.5: Unlocking multimodal understanding across millions of tokens of contextcited by 3,224
  12. 2022Training Compute-Optimal Large Language Modelscited by 3,019
18 more
  1. 2019Fine-Tuning Language Models from Human Preferencescited by 2,150
  2. 2019SciBERT: A Pretrained Language Model for Scientific Textcited by 1,414
  3. 2024OpenAI o1 System Cardcited by 1,244
  4. 2025Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?cited by 982
  5. 2025SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-trainingcited by 672
  6. 2024To CoT or not to CoT? Chain-of-thought helps mainly on math and symbolic reasoningcited by 263
  7. 2023GPT-NER: Named Entity Recognition via Large Language Modelscited by 160
  8. 2025Competitive Programming with Large Reasoning Modelscited by 119
  9. 2024Extract, Define, Canonicalize: An LLM-based Framework for Knowledge Graph Constructioncited by 70
  10. 2023Empirical Study of Zero-Shot NER with ChatGPTcited by 60
  11. 2023Zero-shot Temporal Relation Extraction with ChatGPTcited by 53
  12. 2021Packed Levitated Marker for Entity and Relation Extractioncited by 49
  13. 2020A Frustratingly Easy Approach for Entity and Relation Extractioncited by 37
  14. 2023Revisiting Large Language Models as Zero-shot Relation Extractorscited by 29
  15. 2024SciER: An Entity and Relation Extraction Dataset for Datasets, Methods, and Tasks in Scientific Documentscited by 21
  16. 2024UrbanKGent: A Unified Large Language Model Agent Framework for Urban Knowledge Graph Constructioncited by 15
  17. 2023Joint Entity and Relation Extraction with Span Pruning and Hypergraph Neural Networkscited by 11
  18. 2023PIVOINE: Instruction Tuning for Open-world Information Extractioncited by 10

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026LA-RL: Label-Aware Self-Reflection for Reinforcement Learning in Information Extractioncited by 1
  2. 2026Self-Prompting Small Language Models for Privacy-Sensitive Clinical Information Extractioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025MR-UIE: Multi-Perspective Reasoning with Reinforcement Learning for Universal Information Extraction
  2. 2025From Reasoning to Learning: A Survey on Hypothesis Discovery and Rule Learning with Large Language Models
  3. 2024Flow of Reasoning: Training LLMs for Divergent Reasoning with Minimal Examples
  4. 2025ARCHE: A Novel Task to Evaluate LLMs on Latent Reasoning Chain Extraction
  5. 2026ProjectionBench: Evaluating Scientific Hypothesis Generation in LLMs Under Progressive Information Disclosure
  6. 2023Mastering the Task of Open Information Extraction with Large Language Models and Consistent Reasoning Environment
  7. 2026Learning to Reason Efficiently with A* Post-Training
  8. 2023LLMs for Knowledge Graph Construction and Reasoning: Recent Capabilities and Future Opportunities