#Logical Reasoning

Articles (6)

• 🇺🇸 Draft-and-Prune: Improving the Reliability of Auto-formalization for Logical Reasoning — 19/03/2026 [USA]

  arXiv:2603.17233v1 Announce Type: new Abstract: Auto-formalization (AF) translates natural-language reasoning problems into solver-executable programs, enabling symbolic solvers to perform sound logi...

  Related: #Auto-formalization
• 🇺🇸 Making Implicit Premises Explicit in Logical Understanding of Enthymemes — 09/03/2026 [USA]

  arXiv:2603.06114v1 Announce Type: cross Abstract: Real-world arguments in text and dialogues are normally enthymemes (i.e. some of their premises and/or claims are implicit). Natural language process...

  Related: #Argument Analysis
• 🇺🇸 LLM-Driven Multi-Turn Task-Oriented Dialogue Synthesis for Realistic Reasoning — 02/03/2026 [USA]

  arXiv:2602.23610v1 Announce Type: cross Abstract: The reasoning capability of large language models (LLMs), defined as their ability to analyze, infer, and make decisions based on input information, ...

  Related: #Large Language Models, #Task‑Oriented Dialogue Systems, #Data Benchmarking, #Synthetic Data Generation
• 🇺🇸 Tool Building as a Path to "Superintelligence" — 25/02/2026 [USA]

  arXiv:2602.21061v1 Announce Type: new Abstract: The Diligent Learner framework suggests LLMs can achieve superintelligence via test-time search, provided a sufficient step-success probability $\gamma...

  Related: #Artificial Intelligence, #Superintelligence, #Benchmarking
• 🇺🇸 LogicGraph : Benchmarking Multi-Path Logical Reasoning via Neuro-Symbolic Generation and Verification — 25/02/2026 [USA]

  arXiv:2602.21044v1 Announce Type: new Abstract: Evaluations of large language models (LLMs) primarily emphasize convergent logical reasoning, where success is defined by producing a single correct pr...

  Related: #Artificial Intelligence, #Benchmark Development, #Neuro-Symbolic Systems
• 🇺🇸 Evaluating Robustness of Reasoning Models on Parameterized Logical Problems — 16/02/2026 [USA]

  arXiv:2602.12665v1 Announce Type: new Abstract: Logic provides a controlled testbed for evaluating LLM-based reasoners, yet standard SAT-style benchmarks often conflate surface difficulty (length, wo...

  Related: #AI Evaluation, #Benchmark Development