TIRE: Top-Down Formula Instantiation with Reliable Execution for Physics Problem Synthesis
We introduce TIRE, a framework for Top-Down Formula Instantiation with Reliable Execution that brings knowledge-driven problem construction to large-scale data generation.
ProofPaper ↗
Key points
- Advancing physics reasoning in large language models requires scalable training data that connect physical laws, concrete problem scenarios, and verifiable numerical answers.
- LLM-based synthesis from texts scales efficiently, but model-generated answers and assessments leave supervision vulnerable to hallucinations.
- Top-down instantiation uses reusable physical formulas and their applicability conditions to control the physical relations and numerical configurations of generated problems.
- These results demonstrate the value of generating physics training data from explicit physical knowledge with reliable execution.
Sources (1)
- [1]TIRE: Top-Down Formula Instantiation with Reliable Execution for Physics Problem SynthesisarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 18, 09:38 AM
We introduce TIRE, a framework for Top-Down Formula Instantiation with Reliable Execution that brings knowledge-driven problem construction to large-scale data generation.
Advancing physics reasoning in large language models requires scalable training data that connect physical laws, concrete problem scenarios, and verifiable numerical answers.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 17, 2026Nemotron 3.5 Bolt, Muse Spark Open, and a Fast LFM2.5 24B
- Oct 16, 2026Characterising and Analysing Retrieval-Augmented Generation Techniques
- Oct 16, 2026Where Research Ideation Begins: Forward-Citation Supervision for Actionable Limitation Generation
- Oct 16, 2026Language Models Learn to Improve Language Models
- Oct 16, 2026Tree-Aware Posterior Distillation for Long-Horizon Agentic Reinforcement Learning
- Oct 16, 2026Welcome Beam: a 501B open model you can own, customize, and scale