Deductive Reasoning in Language Models

Follow this topic Ask about this topic

Latest papers

11 papers2 papers in the last four weeks, against 2 the four weeks before. 0.0% of all new papers.

Latent Path Exploration via Potential-Guided Subtrajectory Balance
Oct 29, 2026Xiaoling Zeng, Qi ZhangDeductive Reasoning in Language Models LLM Reasoning
Worcester Polytechnic Institute
Probing for Long-Horizon Deductive Reasoning Capabilities in Language Models with Prolog
Oct 8, 2026Hadeel Al-Negheimish, Jasna Ilieva, Yoon KimDeductive Reasoning in Language Models Long-Horizon LLM Reasoning
King Saud University · Massachusetts Institute of Technology
Learning to Prove, Not Just to Answer: Reinforcement Learning from Formal Verification for Natural-Language Logical Reasoning
Sep 29, 2026Qili Zhang, Qianren Mao, Hanze Cai +11Logical Reasoning RL for Language Model Reasoning
Zhongguancun Laboratory · Beihang University · Nanyang Technological University +1
Clueing up LLMs with Tool-Augmented Deductive Reasoning
Sep 16, 2026Rebecca Ansell, Autumn Toney-WailsMulti-Agent LLM Systems Deductive Reasoning in Language Models
Georgetown University · Syntheos, Corp · UNU-MERIT
Are you Talking Logic to Me? Assessing Language Models Syllogistic Reasoning Capabilities
Jul 22, 2026Hanna Abi Akl, Fabien Gandon, Catherine Faron +1Knowledge Representation Structured Reasoning
Université Côte d’Azur, Inria, CNRS, I3S, Sophia Antipolis, France · Data ScienceTech Institute, Paris, France
QMFOL: Benchmarking Large Language Model Reasoning via Quantifiable Monadic First-Order Logic Test Case Generation
Jun 18, 2026Xinyi Zheng, Ling Shi, Tianlong Yu +3Logical Reasoning LLM Evaluation
Huazhong University of Science and Technology Wuhan, China · Nanyang Technological University Singapore, Singapore · Hubei University Wuhan, China +2
Learning to Reason Efficiently with A* Post-Training
May 23, 2026Andreas Opedal, Francesco Ignazio Re, Abulhair Saparov +3Process Reward Models RL for Language Model Reasoning
αETH Zürich · γPurdue University · βMPI for Intelligent Systems, Tübingen
Rethinking Easy-to-Hard: Limits of Curriculum Learning in Post-Training for Deductive Reasoning
Mar 28, 2026Maximilian Mordig, Andreas Opedal, Weiyang Liu +1Deductive Reasoning in Language Models Curriculum Learning
Max Planck Institute for Intelligent Systems, Tübingen · ETH Zürich · The Chinese University of Hong Kong
How Clued up are LLMs? Evaluating Multi-Step Deductive Reasoning in a Text-Based Game Environment
Mar 17, 2026Rebecca Ansell, Autumn Toney-WailsDeductive Reasoning in Language Models Fine-Tuning Language Models
Department of Computer Science, Georgetown University, Washington, DC, USA · Syntheos, Corp; UNU-Merit, Maastricht, NL
Deductive Logic in Language Models: Horizontal vs Vertical Reasoning
Oct 10, 2025Davide Maltoni, Matteo FerraraTransformer Interpretability Deductive Reasoning in Language Models
Department of Computer Science and Engineering, University of Bologna, Italy
An Incomplete Loop: Deductive, Inductive, and Abductive Reasoning in Language Models
Apr 3, 2024Emmy Liu, Graham Neubig, Jacob AndreasDeductive Reasoning in Language Models LLM Prompting
Language Technologies Institute Carnegie Mellon University · CSAIL Massachusetts Institute of Technology

Oct 29, 2026J/K move · Enter open · S save

Latent Path Exploration via Potential-Guided Subtrajectory Balance

Xiaoling Zeng, Qi Zhang

Worcester Polytechnic Institute

Deductive Reasoning in Language Models LLM Reasoning Continuous Latent Reasoning Numerical Reasoning in Language Models

Open paper AskPDF arXiv

Continuous latent reasoning enables language models to perform intermediate computation without generating explicit reasoning traces. Advances in stochastic latent reasoning allow multiple paths to be sampled in continuous space, but stochasticity alone does not ensure effective exploration. A key challenge is guiding stochastic transitions between intermediate latent states toward useful reasoning paths. We introduce SubLAT, which couples a relative answer potential with Subtrajectory Balance to learn stochastic latent exploration. The relative potential evaluates the answer support of sampled states against replacement states, with a correctness signal at the terminal state. Subtrajectory Balance relates potential differences to forward and backward transition densities across subtrajectories of different lengths. The model is jointly trained with answer supervision, CoT supervision, and representation distillation. Experiments on arithmetic and deductive reasoning show increasing solution coverage with larger sampling budgets across in-domain and out-of-domain benchmarks, as well as improvements on several structural OOD benchmarks. Trajectory analyses reveal progressive divergence among sampled latent paths and higher mean relative potentials for correct trajectories across arithmetic benchmarks.