# IGDA: Interactive Graph Discovery through Large Language Model Agents

arXiv: https://arxiv.org/abs/2502.17189 · HTML: https://mlsanity.com/paper/2502.17189/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Web, search, and research agents, one of 16 directions in LLM agents and tool use.
- Builds on 17 of its 30 references that are in the map.
- First posted on arXiv 24 Feb 2025.

## Placed in

- LLM agents and tool use / [Web, search, and research agents](https://mlsanity.com/agents/web-search-and-research-agents/)

## Builds on (most cited first, 17 of 17)

- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Large Language Models as Optimizers](https://mlsanity.com/paper/2309.03409/) (2023)
- [DAGs with NO TEARS: Continuous Optimization for Structure Learning](https://mlsanity.com/paper/1803.01422/) (2018)
- [Causal Reasoning and Large Language Models: Opening a New Frontier for Causality](https://mlsanity.com/paper/2305.00050/) (2023)
- [DAG-GNN: DAG Structure Learning with Graph Neural Networks](https://mlsanity.com/paper/1904.10098/) (2019)
- [Causal Inference and Causal Explanation with Background Knowledge](https://mlsanity.com/paper/1302.4972/) (2013)
- [The Impact of Large Language Models on Scientific Discovery: a Preliminary Study using GPT-4](https://mlsanity.com/paper/2311.07361/) (2023)
- [GLoRe: When, Where, and How to Improve LLM Reasoning via Global and Local Refinements](https://mlsanity.com/paper/2402.10963/) (2024)
- [Efficient Neural Causal Discovery without Acyclicity Constraints](https://mlsanity.com/paper/2107.10483/) (2021)
- [BioDiscoveryAgent: An AI Agent for Designing Genetic Perturbation Experiments](https://mlsanity.com/paper/2405.17631/) (2024)
- [Causal Order: The Key to Leveraging Imperfect Experts in Causal Inference](https://mlsanity.com/paper/2310.15117/) (2023)
- [Learning Neural Causal Models with Active Interventions](https://mlsanity.com/paper/2109.02429/) (2021)
- [Neuropathic Pain Diagnosis Simulator for Causal Discovery Algorithm Evaluation](https://mlsanity.com/paper/1906.01732/) (2019)
- [CausalBench: A Large-scale Benchmark for Network Inference from Single-cell Perturbation Data](https://mlsanity.com/paper/2210.17283/) (2022)
- [Passive learning of active causal strategies in agents and language models](https://mlsanity.com/paper/2305.16183/) (2023)
- [CausalBN-Bench: A Comprehensive Benchmark for Causal Learning Capability of LLMs](https://mlsanity.com/paper/2404.06349/) (2024)
- [CLEAR: Can Language Models Really Understand Causal Graphs?](https://mlsanity.com/paper/2406.16605/) (2024)

## Built on by (most cited first, 3 of 3)

- [Graphs Meet AI Agents: Taxonomy, Progress, and Future Opportunities](https://mlsanity.com/paper/2506.18019/) (2025)
- [Think Locally, Explain Globally: Graph-Guided LLM Investigations via Local Reasoning and Belief Propagation](https://mlsanity.com/paper/2601.17915/) (2026)
- [Improving constraint-based discovery with robust propagation and reliable LLM priors](https://mlsanity.com/paper/2509.23570/) (2025)

## Abstract (arXiv)

Large language models (LLMs) have emerged as a powerful method for discovery. Instead of utilizing numerical data, LLMs utilize associated variable textitsemantic metadata to predict variable relationships. Simultaneously, LLMs demonstrate impressive abilities to act as black-box optimizers when given an objective f and sequence of trials. We study LLMs at the intersection of these two capabilities by applying LLMs to the task of textitinteractive graph discovery: given a ground truth graph G^* capturing variable relationships and a budget of I edge experiments over R rounds, minimize the distance between the predicted graph hatGR and G^* at the end of the R-th round. To solve this task we propose IGDA, a LLM-based pipeline incorporating two key components: 1) an LLM uncertainty-driven method for edge experiment selection 2) a local graph update strategy utilizing binary feedback from experiments to improve predictions for unselected neighboring edges. Experiments on eight different real-world graphs show our approach often outperforms all baselines including a state-of-the-art numerical method for interactive graph discovery. Further, we conduct a rigorous series of ablations dissecting the impact of each pipeline component. Finally, to assess the impact of memorization, we apply our interactive graph discovery strategy to a complex, new (as of July 2024) causal graph on protein transcription factors, finding strong performance in a setting where memorization is impossible. Overall, our results show IGDA to be a powerful method for graph discovery complementary to existing numerically driven approaches.