All research directions

Machine Reasoning

Plan-guided learning, structured inference, and reasoning capabilities in large language models.

To equip AI systems with structured reasoning — planning, multi-step inference, and grounded decision-making beyond pattern matching.

Overview

This research explores how large models perform structured reasoning and planning. Work includes plan-guided tuning, cross-lingual topic modeling with LLMs, and methods that improve logical and causal inference in foundation models.

Key objectives

  • Develop plan-guided and structured learning methods
  • Improve multi-step reasoning in language models
  • Integrate causal and logical structure into inference
  • Enable grounded decision-making in complex tasks

Key topics

  • Plan-guided learning and tuning
  • Structured and multi-step inference
  • Cross-lingual and topic reasoning with LLMs
  • Causal and logical reasoning

Papers in this direction

  • 2026

    LLM-XTM: Enhancing Cross-Lingual Topic Models with Large Language Models

    Xuan, MC, Nguyen, TP, Van, LN, Sang, DV, Diep, NTN, Le, T

    arXiv preprint arXiv:2605.03299

  • 2026

    Causal-aware Anomaly Detection for Tabular Data

    Nguyen, D, Nguyen, TAH, Le, TD, Venkatesh, S, Le, T, Gupta, S

    International Conference on Machine Learning (ICML)