Large Language Models (LLMs) like closed weights ones GPT-3.5/4, Claude, Gemini or open weights ones like LLaMa 2/3, Mistral, Mixtral, and more recent ones Dbrx or Command R+ are often described as being instances of foundation models - that is, models that transfer strongly across various tasks and conditions in few-show or zero-shot manner, while exhibiting scaling laws that predict function improvement when increasing the pre-training scale. These claims of excelling in different functions and tasks rely on measurements taken across various sets of standardized benchmarks showing high scores for such models. We demonstrate here a dramatic breakdown of function and reasoning capabilities of state-of-the-art models trained at the largest available scales which claim strong function, using a simple, short, conventional problem formulated in concise natural language, easily solvable by humans. The breakdown is dramatic, as models also express strong overconfidence in their wrong solutions, while providing often non-sensical "reasoning"-like explanations akin to confabulations to justify and backup the validity of their clearly failed responses, making them sound plausible. Various standard interventions in an attempt to get the right solution, like various type of enhanced prompting, or urging the models to reconsider the wrong solutions again by multi step re-evaluation, fail. We take these initial observations to the scientific and technological community to stimulate urgent re-assessment of the claimed capabilities of current generation of LLMs, Such re-assessment also requires common action to create standardized benchmarks that would allow proper detection of such basic reasoning deficits that obviously manage to remain undiscovered by current state-of-the-art evaluation procedures and benchmarks.
@article{nezhurina2024alice,
title={Alice in Wonderland: Simple Tasks Showing Complete Reasoning Breakdown in State-Of-the-Art Large Language Models},
author={Marianna Nezhurina and Lucia Cipolina-Kun and Mehdi Cherti and Jenia Jitsev},
year={2024},
journal={arXiv preprint arXiv:2406.02061},
eprint={2406.02061},
archivePrefix={arXiv},
primaryClass={cs.LG}
}We would like to express gratitude to all the people who are working on making code, models and data publicly available, advancing community based research and making research more reproducible. Specifically, we would like to thank all the members of the LAION Discord server community and Open-Ψ (Open-Sci) Collective for providing fruitful ground for scientific exchange and open-source development. Marianna Nezhurina acknowledges funding by the Federal Ministry of Education and Research of Germany under grant no. 01IS22094B WestAI - AI Service Center West. Lucia Cipolina-Kun acknowledges the Helmholtz Information & Data Science Academy (HIDA) for providing financial support enabling a short-term research stay at Juelich Supercomputing Center (JSC), Research Center Juelich (FZJ) to conduct research on foundation models.