mlsanityData to 14 Oct 2026

Training a Scientific Reasoning Model for Chemistry

arXiv 2506.17238 ↗First posted 4 Jun 2025cs.LGcited by 56

On the map: LLM reasoning → Clinical and biomedical applications

Abstract (arXiv)

Reasoning models are large language models that emit a long chain-of-thought before answering, providing both higher accuracy and explicit reasoning for their response. A major question has been whether language model reasoning generalizes beyond mathematics, programming, and logic, where most previous work has focused. We demonstrate that reasoning models can be post-trained for chemistry without additional domain pretraining, and require substantially less data compared to contemporary domain-specific models. We report ether0, a 24B parameter LLM (based on Mistral-Small-24B) that can reason in natural language and respond with chemical structures. This reasoning model was trained with reinforcement learning on 640,730 experimentally-grounded chemistry problems across 375 tasks ranging from synthesizability, to blood-brain barrier permeability, to human receptor activity, to scent. Our model exceeds general-purpose chemistry models, frontier models, and human experts on molecular design tasks. It is also more data efficient relative to specialized models. We anticipate that this method can be applied to train data-efficient language models specialized for tasks across a wide variety of scientific domains.

Key facts. Cited by 56 papers in the map (data through 14 Oct 2026). Placed in Clinical and biomedical applications, one of 17 directions in LLM reasoning. Builds on 38 of its 116 references that are in the map. First posted on arXiv 4 Jun 2025.

Builds on

38 of 116 references.

Closest ancestors · 5

Same direction, newest first

  1. 2025BioReason: Incentivizing Multimodal Biological Reasoning within a DNA-LLM Model29 May
  2. 2025Med-R1: Reinforcement Learning for Generalizable Medical Reasoning in Vision-Language Models18 Mar
  3. 2025Chemical reasoning in LLMs unlocks strategy-aware synthesis planning and reaction mechanism elucidation11 Mar
  4. 2025O1 Replication Journey -- Part 3: Inference-time Scaling for Medical Reasoning11 Jan
  5. 2024Structural Reasoning Improves Molecular Understanding of LLM8 Oct

Other references · 33

  1. 2017Proximal Policy Optimization Algorithmscited by 15,215
  2. 2024The Llama 3 Herd of Modelscited by 14,715
  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. 2025DAPO: An Open-Source LLM Reinforcement Learning System at Scalecited by 2,810
  8. 2025s1: Simple test-time scalingcited by 1,341
  9. 2024OpenAI o1 System Cardcited by 1,244
  10. 2025Kimi k1.5: Scaling Reinforcement Learning with LLMscited by 1,210
  11. 2024Large Language Monkeys: Scaling Inference Compute with Repeated Samplingcited by 991
  12. 2024DeepSeek-V3 Technical Reportcited by 811
21 more
  1. 2024Phi-4 Technical Reportcited by 711
  2. 2025Humanity's Last Examcited by 491
  3. 2025Cognitive Behaviors that Enable Self-Improving Reasoners, or, Four Habits of Highly Effective STaRscited by 377
  4. 2025Reinforcement Learning for Reasoning in Large Language Models with One Training Examplecited by 295
  5. 2024Teaching Large Language Models to Reason with Reinforcement Learningcited by 160
  6. 2025Llama-Nemotron: Efficient Reasoning Modelscited by 158
  7. 2024O1 Replication Journey: A Strategic Progress Report -- Part 1cited by 143
  8. 2025Nemotron-H: A Family of Accurate and Efficient Hybrid Mamba-Transformer Modelscited by 100
  9. 2024The Surprising Effectiveness of Test-Time Training for Few-Shot Learningcited by 96
  10. 2024O1 Replication Journey -- Part 2: Surpassing O1-preview through Simple Distillation, Big Progress or Bitter Lesson?cited by 92
  11. 2025InverseBench: Benchmarking Plug-and-Play Diffusion Priors for Inverse Problems in Physical Sciencescited by 68
  12. 2024Interpretable Contrastive Monte Carlo Tree Search Reasoningcited by 64
  13. 2025Nemotron-CrossThink: Scaling Self-Learning beyond Math Reasoningcited by 31
  14. 2024Aviary: training language agents on challenging scientific taskscited by 26
  15. 2024MolecularGPT: Open Large Language Model (LLM) for Few-Shot Molecular Property Predictioncited by 24
  16. 2025OmniScience: A Domain-Specialized LLM for Scientific Reasoning and Discoverycited by 20
  17. 2024SciKnowEval: Evaluating Multi-level Scientific Knowledge of Large Language Modelscited by 19
  18. 2024SciSafeEval: A Comprehensive Benchmark for Safety Alignment of Large Language Models in Scientific Taskscited by 18
  19. 2024DrugLLM: Open Large Language Model for Few-shot Molecule Generationcited by 15
  20. 2023In-Context Learning for Few-Shot Molecular Property Predictioncited by 8
  21. 2024LLM-based MOFs Synthesis Condition Extraction using Few-Shot Demonstrationscited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 17; 2026: 39.

56 papers.

Sort:

By area:

  1. 2025A Survey of Reinforcement Learning for Large Reasoning Modelscited by 180
  2. 2025ChemDFM-R: A Chemical Reasoning LLM Enhanced with Atomized Chemical Knowledgecited by 20
  3. 2025MolReasoner: Toward Effective and Interpretable Reasoning for Molecular LLMscited by 12
  4. 2025Autonomous Agents for Scientific Discovery: Orchestrating Scientists, Language, Code, and Physicscited by 11
  5. 2025Synthelite: Chemist-aligned and feasibility-aware synthesis planning with LLMscited by 10
  6. 2026MolecularIQ: Characterizing Chemical Reasoning Capabilities Through Symbolic Verification on Molecular Graphscited by 8
  7. 2026LatentChem: From Textual CoT to Latent Thinking in Chemical Reasoningcited by 8
  8. 2025Quantum Verifiable Rewards for Post-Training Qiskit Code Assistantcited by 6
  9. 2026Agentic reinforcement learning empowers next-generation chemical language models for molecular design and synthesiscited by 6
  10. 2026El Agente Estructural: An Artificially Intelligent Molecular Editorcited by 5
  11. 2025OwkinZero: Accelerating Biological Discovery with AIcited by 4
  12. 2025oMeBench: Towards Robust Benchmarking of LLMs in Organic Mechanism Elucidation and Reasoningcited by 4
38 more
  1. 2025Atom-anchored LLMs speak Chemistry: A Retrosynthesis Demonstrationcited by 4
  2. 2026Evaluating the Progression of Large Language Model Capabilities for Small-Molecule Drug Designcited by 4
  3. 2025SynthStrategy: Extracting and Formalizing Latent Strategic Insights from LLMs in Organic Chemistrycited by 3
  4. 2025BioMedGPT-Mol: Multi-task Learning for Molecular Understanding and Generationcited by 3
  5. 2025MiST: Understanding the Role of Mid-Stage Scientific Training in Developing Chemical Reasoning Modelscited by 3
  6. 2026Segment-Level Coherence for Robust Harmful Intent Probing in LLMscited by 3
  7. 2026Molecular Representations for Large Language Modelscited by 3
  8. 2026Chemical Chain-of-Thought Functions as a Hallucination-Prone Molecular Scratchpadcited by 3
  9. 2025Knowledge-Augmented Long-CoT Generation for Complex Biomolecular Reasoningcited by 2
  10. 2025Generalization of RLVR Using Causal Reasoning as a Testbedcited by 2
  11. 2026DrugR: Optimizing Molecular Drugs through LLM-based Explicit Reasoningcited by 2
  12. 2026ChemVLR: Prioritizing Reasoning in Perception for Chemical Vision-Language Understandingcited by 2
  13. 2026From Answers to States: Verifiable Process-Level Evaluation of Chemical Reasoning in Large Language Modelscited by 2
  14. 2026ChemDIRT: A Diversified Instruction, Representation, and Task Benchmark for Robust Chemistry-LLM Evaluationcited by 2
  15. 2025RTMol: Rethinking Molecule-text Alignment in a Round-trip Viewcited by 1
  16. 2026D3-Gym: Constructing Real-World Verifiable Environments for Data-Driven Discoverycited by 1
  17. 2026From Knowledge to Action: Outcomes of the 2025 Large Language Model (LLM) Hackathon for Applications in Materials Science and Chemistrycited by 1
  18. 2026How Post-Training Shapes Biological Reasoning Modelscited by 1
  19. 2026Accurate, Interdisciplinary and Transparent Structure-property Understanding with Deep Native Structural Reasoningcited by 1
  20. 2026Learning Mechanistic Reasoning for Chemical Reactions with Large Language Modelscited by 1
  21. 2025OPENXRD: A Comprehensive Benchmark Framework for LLM/MLLM XRD Question Answeringcited by 0
  22. 2025A Scientific Reasoning Model for Organic Synthesis Procedure Generationcited by 0
  23. 2026Teaching and Evaluating LLMs to Reason About Polymer Design Related Taskscited by 0
  24. 2026Beyond Learning on Molecules by Weakly Supervising on Moleculescited by 0
  25. 2026MMAI Gym for Science: Training Liquid Foundation Models for Drug Discoverycited by 0
  26. 2026RetroReasoner: A Reasoning LLM for Strategic Retrosynthesis Predictioncited by 0
  27. 2026Protein Design with Agent Rosetta: A Case Study for Specialized Scientific Agentscited by 0
  28. 2026MolRGen: A Training and Evaluation Setting for De Novo Molecular Generation with Reasonning Modelscited by 0
  29. 2026Deciphering Scientific Reasoning Steps from Outcome Data for Molecule Optimizationcited by 0
  30. 2026PubSwap: Public-Data Off-Policy Coordination for Federated RLVRcited by 0
  31. 2026RosettaSearch: Multi-Objective Inference-Time Search for Protein Sequence Designcited by 0
  32. 2026MolDeTox: Evaluating Language Model's Stepwise Fragment Editing for Molecular Detoxificationcited by 0
  33. 2026Rethinking Molecular Text Representations for LLMs: An Empirical Studycited by 0
  34. 2026Mat-Pref: Verifiable-Reward Training Improves Compositional Reasoning in Inorganic Materialscited by 0
  35. 2026Agents in the Wild: Where Research Meets Deploymentcited by 0
  36. 2026From Execution to Capability: Scientific Experience Consolidation via Procedural Knowledge Synthesiscited by 0
  37. 2026MolGVR: A Chemistry-Grounded Framework for Text-to-Molecule Generationcited by 0
  38. 2026Molecular LLM Agents: From Architectural Design to Scientific Autonomycited by 0

50 most cited of 56 All 56 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025ChemATP: A Training-Free Chemical Reasoning Framework for Large Language Models
  2. 2025Assessing the Chemical Intelligence of Large Language Models
  3. 2025SUPERChem: A Multimodal Reasoning Benchmark in Chemistry
  4. 2025ChemAU: Harness the Reasoning of LLMs in Chemical Research with Adaptive Uncertainty Estimation
  5. 2026ChemPro: A Progressive Chemistry Benchmark for Large Language Models
  6. 2025ChemAgent: Self-updating Library in Large Language Models Improves Chemical Reasoning
  7. 2025Beyond Chemical QA: Evaluating LLM's Chemical Reasoning with Modular Chemical Operations
  8. 2024Are large language models superhuman chemists?