@article{reasonbenchbenchmarkingtheinstabilityofl, title = {ReasonBENCH: Benchmarking the (In)Stability of LLM Reasoning}, author = {Nearchos Potamitis and Vansh Ramani and Har Ashish Arora and Dhairya Kuchhal and Lars Klein and Akhil Arora}, year = {2025}, eprint = {2512.07795}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2512.07795}, }