Difference between revisions of "Science Agents"

From GISAXS

Jump to: navigation, search

Latest revision as of 18:52, 28 January 2026

Contents

1 AI Use-cases for Science
2 Science Benchmarks
3 Science Agents
4 AI Science Systems
5 Impact of AI in Science
6 Related Tools
7 Science Datasets
8 Genuine Discoveries
9 See Also

AI Use-cases for Science

Literature

alphaXiv | Explore: Understand arXiv papers

LLM extract data from papers

2024-14: From text to insight: large language models for chemical data extraction

AI finding links in literature

(Pre) Generate Articles

2022-12: Re3: Generating Longer Stories With Recursive Reprompting and Revision
2023-03: English essays: Artificial intelligence (AI) technology in OpenAI ChatGPT application: A review of ChatGPT in writing English essay
2023-01: Journalism: Collaborating With ChatGPT: Considering the Implications of Generative Artificial Intelligence for Journalism and Media Education
2023-07: Science writing: Artificial intelligence in scientific writing: a friend or a foe?
2024-02: Wikipedia style: Assisting in Writing Wikipedia-like Articles From Scratch with Large Language Models
2024-02: LongWriter: Unleashing 10,000+ Word Generation from Long Context LLMs (code)
2024-08: Scientific papers: The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery
2024-09: PaperQA2: Language Models Achieve Superhuman Synthesis of Scientific Knowledge (𝕏 post, code)
2025-03: Reasoning to Learn from Latent Thoughts
2025-03: WikiAutoGen: Towards Multi-Modal Wikipedia-Style Article Generation
2025-04: Sleep-time Compute: Beyond Inference Scaling at Test-time

Explanation

2025-02: TheoremExplainAgent: Towards Multimodal Explanations for LLM Theorem Understanding (preprint)
2025-04: Do Two AI Scientists Agree?

Autonomous Ideation

Adapting LLMs to Science

AI/LLM Control of Scientific Instruments/Facilities

AI/ML Methods tailored to Science

Science Foundation Models

2025-08: Intern-S1: A Scientific Multimodal Foundation Model
2025-11: A foundation model for atomistic materials chemistry
2025-11: Walrus: A Cross-Domain Foundation Model for Continuum Dynamics
2026-01: Deep contrastive learning enables genome-wide virtual screening

Regression (Data Fitting)

2024-06: Connecting the Dots: LLMs can Infer and Verbalize Latent Structure from Disparate Training Data: training on (x,y) pairs enables inferring underlying function (define it in code, invert it, compose it)
2024-12: OmniPred: Language Models as Universal Regressors

Tabular Classification/Regression

2025-01: Accurate predictions on small data with a tabular foundation model (code)

Symbolic Regression

2024-09: Symbolic Regression with a Learned Concept Library

Literature Discovery

Commercial

Sakana AI
Cusp AI: Materials/AI
Lila AI: Life sciences
Radical AI: Material simulation/design
Autoscience (Carl)
Periodic Labs
Edison Scientific (drug discovery, spinoff from FutureHouse)

Bio

AI/ML Methods in Science

2025-07: Synthetic Scientific Image Generation with VAE, GAN, and Diffusion Model Architectures

Imaging

2025-05: Behind the Noise: Conformal Quantile Regression Reveals Emergent Representations (blog: Behind the Noise)

Materials

Chemistry

2025-01: Large language models for reticular chemistry
2025-02: Image-based generation for molecule design with SketchMol
2025-02: Large language models for scientific discovery in molecular property prediction
2025-03: Vant AI Neo-1: atomistic foundation model (small molecules, proteins, etc.)
2025-04: Compositional Flows for 3D Molecule and Synthesis Pathway Co-design
2025-07: General purpose models for the chemical sciences
2025-11: ChemTorch: A Deep Learning Framework for Benchmarking and Developing Chemical Reaction Property Prediction Models

Biology

2018: AlphaFold
2021-07: AlphaFold 2
2024-05: AlphaFold 3
2023-03: Evolutionary-scale prediction of atomic-level protein structure with a language model (ESMFold)
2023-11: Illuminating protein space with a programmable generative model
2024-11: Sequence modeling and design from molecular to genome scale with Evo (Evo)
2025-01: Targeting protein–ligand neosurfaces with a generalizable deep learning tool (Chroma)
2025-01: Simulating 500 million years of evolution with a language model (ESM 3 model)
2025-02: Genome modeling and design across all domains of life with Evo 2
2025-02: Exploring the structural changes driving protein function with BioEmu-1
2025-02: Protein Large Language Models: A Comprehensive Survey
2025-03: Vant AI Neo-1: atomistic foundation model (small molecules, proteins, etc.)
2025-03: Lyra: An Efficient and Expressive Subquadratic Architecture for Modeling Biological Sequences
2025-08: RosettaFold 3: Accelerating Biomolecular Modeling with AtomWorks and RF3
2025-09: Generative design of novel bacteriophages with genome language models
2025-10: Strengthening nucleic acid biosecurity screening against generative protein design tools
2026-01: Advancing regulatory variant effect prediction with AlphaGenome

Medicine

See: AI_Agents#Medicine

Successes

2025-02: Site-Decorated Model for Unconventional Frustrated Magnets: Ultranarrow Phase Crossover and Spin Reversal Transition

AI/ML Methods co-opted for Science

Mechanistic Interpretability

Train large model on science data. Then apply mechanistic interpretability (e.g. sparse autoencoders, SAE) to the feature/activation space.

Mechanistic interpretability for protein language models (visualizer, code, SAE)
Markov Bio: Through a Glass Darkly: Mechanistic Interpretability as the Bridge to End-to-End Biology (quick description, background info on recent bio progress)
2023-01: Tracr: Compiled Transformers as a Laboratory for Interpretability (code)
2024-10: An X-Ray Is Worth 15 Features: Sparse Autoencoders for Interpretable Radiology Report Generation
2024-12: Towards scientific discovery with dictionary learning: Extracting biological concepts from microscopy foundation models
2024-12: InterPLM: Discovering Interpretable Features in Protein Language Models via Sparse Autoencoders
2025-01: Insights on Galaxy Evolution from Interpretable Sparse Feature Networks
2025-02: From Mechanistic Interpretability to Mechanistic Biology: Training, Evaluating, and Interpreting Sparse Autoencoders on Protein Language Models
2025-02: Interpreting Evo 2: Arc Institute's Next-Generation Genomic Foundation Model
2026-01: Using Interpretability to Identify a Novel Class of Alzheimer's Biomarkers

Uncertainty

Science Benchmarks

2024-07: SciCode: A Research Coding Benchmark Curated by Scientists (project)
2024-11: AidanBench: Evaluating Novel Idea Generation on Open-Ended Questions (code)
2024-12: LiveIdeaBench: Evaluating LLMs' Scientific Creativity and Idea Generation with Minimal Context
2025-01: Humanity's Last Exam
ScienceAgentBench
2025-02: EAIRA: Establishing a Methodology for Evaluating AI Models as Scientific Research Assistants
2025-03: BixBench: Novel hypotheses (accept/reject)
2025-04: Google: Evaluating progress of LLMs on scientific problem-solving

Science Agents

Reviews

Challenges

2026-01: Why LLMs Aren't Scientists Yet: Lessons from Four Autonomous Research Attempts

Specific

2024-01-13: ORGANA: A Robotic Assistant for Automated Chemistry Experimentation and Characterization (video)
2024-06-19: LLMatDesign: Autonomous Materials Discovery with Large Language Models
2024-08-12: Sakana AI: AI Scientist; The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery (code)
2024-09-09: SciAgents: Automating scientific discovery through multi-agent intelligent graph reasoning (code)
2024-09-11: PaperQA2: Language Models Achieve Superhuman Synthesis of Scientific Knowledge (𝕏 post, code)
2024-10-17: Rapid and Automated Alloy Design with Graph Neural Network-Powered LLM-Driven Multi-Agent Systems
2024-10-28: Large Language Model-Guided Prediction Toward Quantum Materials Synthesis
2024-12-06: The Virtual Lab: AI Agents Design New SARS-CoV-2 Nanobodies with Experimental Validation (writeup: Virtual lab powered by ‘AI scientists’ super-charges biomedical research: Could human–AI collaborations be the future of interdisciplinary studies?)
2024-12-30: Aviary: training language agents on challenging scientific tasks
See also: AI Agents > Deep Research
2025-04-08: Sakana: The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search (code)
2025-07: DREAMS: Density Functional Theory Based Research Engine for Agentic Materials Simulation
2025-11: Kosmos: An AI Scientist for Autonomous Discovery
2025-11: SciAgent: A Unified Multi-Agent System for Generalistic Scientific Reasoning

Science Multi-Agent Setups

2025-01: Agent Laboratory: Using LLM Agents as Research Assistants
2025-04: Coordinated AI agents for advancing healthcare (pdf)

AI Science Systems

2025-01: Dolphin: Closed-loop Open-ended Auto-research through Thinking, Practice, and Feedback
2025-01: Hypothesis Generation for Materials Discovery and Design Using Goal-Driven and Constraint-Guided LLM Agents
2025-02: Towards an AI co-scientist (Google blog post: Accelerating scientific breakthroughs with an AI co-scientist)
2025-06: The Discovery Engine
- 2025-07: Benchmarking the Discovery Engine (blog)
2025-07: Autonomous Scientific Discovery Through Hierarchical AI Scientist Systems
2025-12: Probing Scientific General Intelligence of LLMs with Scientist-Aligned Workflows
2026-01: SciSciGPT: advancing human–AI collaboration in the science of science

Inorganic Materials Discovery

2023-11: Scaling deep learning for materials discovery
2023-11: An autonomous laboratory for the accelerated synthesis of novel materials
2024-09: HoneyComb: A Flexible LLM-Based Agent System for Materials Science
2024-10: Open Materials 2024 (OMat24) Inorganic Materials Dataset and Models (code, datasets, checkpoints, blogpost)
2025-01: A generative model for inorganic materials design
2025-04: System of Agentic AI for the Discovery of Metal-Organic Frameworks
2025-05: The Open Molecules 2025 (OMol25) Dataset, Evaluations, and Models

Materials Characterization

2025-08: Operationalizing Serendipity: Multi-Agent AI Workflows for Enhanced Materials Characterization with Theory-in-the-Loop

Chemistry

2023-12: Autonomous chemical research with large language models (Coscientist)
2024-09: PNNL ChemAIst V0.2
2024-11: An automatic end-to-end chemical synthesis development platform powered by large language models
2025-06: Training a Scientific Reasoning Model for Chemistry
2025-06: ChemGraph: An Agentic Framework for Computational Chemistry Workflows (code)

Bio

2025-07: BioMARS: A Multi-Agent Robotic System for Autonomous Biological Experiments

Physics

2025-12: PhysMaster: Building an Autonomous AI Physicist for Theoretical and Computational Physics Research

LLMs Optimized for Science

2022-11: Galactica: A Large Language Model for Science
2024-12: Crystal structure generation with autoregressive large language modeling
2025-02: MatterChat: A Multi-Modal LLM for Material Science
2025-03: OmniScience: A Domain-Specialized LLM for Scientific Reasoning and Discovery
2025-03: Google TxGemma (2B, 9B, 27B): drug development

Impact of AI in Science

2024-11: ~~Artificial Intelligence, Scientific Discovery, and Product Innovation~~
- 2025-05: Retraction: Assuring an accurate research record
2025-02: Transforming Science with Large Language Models: A Survey on AI-assisted Scientific Discovery, Experimentation, Content Generation, and Evaluation

Related Tools

Literature Search

Data Visualization

2024-10: Microsoft Data Formulator: Create Rich Visualization with AI iteratively (video, code)
Julius AI: Analyze your data with computational AI

Generative

2025-03: StarVector 1B, 8B: text or image to SVG

Chemistry

2025-03: Rxn-INSIGHT: fast chemical reaction analysis using bond-electron matrices (docs)

Science Datasets

Genuine Discoveries

2025-11: Early science acceleration experiments with GPT-5
2025-12: AI can obviously create new knowledge - But maybe not new concepts
Math:
- 2023-07: Faster sorting algorithms discovered using deep reinforcement learning
- 2025-06: AlphaEvolve: A coding agent for scientific and algorithmic discovery
- 2025-11: Mathematical exploration and discovery at scale
- 2025-11: Olympiad-level formal mathematical reasoning with reinforcement learning
- 2025-12: Extremal descendant integrals on moduli spaces of curves: An inequality discovered and proved in collaboration with AI
- AI Solving Erdős Problems:
  - 2026-01: Erdős Problem #728 and #729 solved by Aristotle using ChatGPT 5.2 Pro
  - 2026-01: Erdős Problem #397 solved by Neel Somani using ChatGPT 5.2 Pro
  - 2026-01: Erdős Problem #205 solved by Aristotle using ChatGPT 5.2 Pro
  - 2026-01: Erdős Problem #281 solved by Neel Somani using ChatGPT 5.2 Pro
- 2026-01: The motivic class of the space of genus 0 maps to the flag variety
Physics assistance:
- 2025-03: Exact solution of the frustrated Potts model with next-nearest-neighbor interactions in one dimension via AI bootstrapping (Weiguo Yin)
- 2025-12: Relativistic covariance and nonlinear quantum mechanics: Tomonaga-Schwinger analysis
  - Steve Hsu, Theoretical Physics with Generative AI
Literature exploration:
- 2025-11: Kosmos: An AI Scientist for Autonomous Discovery (Edison)
Bio design:
- 2023-07: De novo design of protein structure and function with RFdiffusion
- 2025-11: Atomically accurate de novo design of antibodies with RFdiffusion
- 2025-11: AlphaFold: Five years of impact
- 2026-01: Using Interpretability to Identify a Novel Class of Alzheimer's Biomarkers
Material Discovery:
- 2023-11: Scaling deep learning for materials discovery

See Also

AI agents
Nanobot.chat: Intelligent AI for the labnetwork @ mtl.mit.edu forum

Retrieved from "http://gisaxs.com/index.php?title=Science_Agents&oldid=8594"