AInstein: Can LLMs Solve Research Problems From Parametric Memory Alone?

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Mishra, Shambhavi, Sahu, Gaurav, Pedersoli, Marco, Charlin, Laurent, Dolz, Jose, Pal, Christopher
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918471373160448
author Mishra, Shambhavi
Sahu, Gaurav
Pedersoli, Marco
Charlin, Laurent
Dolz, Jose
Pal, Christopher
author_facet Mishra, Shambhavi
Sahu, Gaurav
Pedersoli, Marco
Charlin, Laurent
Dolz, Jose
Pal, Christopher
contents Can large language models solve AI research problems using only their parametric knowledge, without fine-tuning, retrieval, or other external aids? We introduce AInstein, a framework for testing whether LLM agents can generate and refine solutions to research problems through iterative critique loops. A blind study with 20 domain experts on held-out ICLR 2026 problems validates our automated metrics, which we then scale to 1,214 ICLR 2025 papers using an LLM-as-a-judge paradigm. Two metrics capture complementary aspects of performance: Success Rate (does the solution address the problem?) and Rediscovery (does it match the published approach?). LLMs succeed on over 70% of problems, yet strictly rediscover the published solution less than 19% of the time, suggesting genuine problem-solving rather than associative recall. However, this ability has clear limits: models handle familiar methodological territory well but fail when solutions require cross-domain analogical transfer, a pattern we call the parametric knowledge boundary. On the ResearchPlanGen benchmark (2,645 problems), our training-free iterative refinement strategy matches RL finetuning, and a criteria-coverage analysis pins down the ceiling of what test-time refinement alone can achieve. Together, these findings map both the capabilities and the limits of LLMs as autonomous scientific problem-solvers.
format Preprint
id arxiv_https___arxiv_org_abs_2510_05432
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle AInstein: Can LLMs Solve Research Problems From Parametric Memory Alone?
Mishra, Shambhavi
Sahu, Gaurav
Pedersoli, Marco
Charlin, Laurent
Dolz, Jose
Pal, Christopher
Artificial Intelligence
Can large language models solve AI research problems using only their parametric knowledge, without fine-tuning, retrieval, or other external aids? We introduce AInstein, a framework for testing whether LLM agents can generate and refine solutions to research problems through iterative critique loops. A blind study with 20 domain experts on held-out ICLR 2026 problems validates our automated metrics, which we then scale to 1,214 ICLR 2025 papers using an LLM-as-a-judge paradigm. Two metrics capture complementary aspects of performance: Success Rate (does the solution address the problem?) and Rediscovery (does it match the published approach?). LLMs succeed on over 70% of problems, yet strictly rediscover the published solution less than 19% of the time, suggesting genuine problem-solving rather than associative recall. However, this ability has clear limits: models handle familiar methodological territory well but fail when solutions require cross-domain analogical transfer, a pattern we call the parametric knowledge boundary. On the ResearchPlanGen benchmark (2,645 problems), our training-free iterative refinement strategy matches RL finetuning, and a criteria-coverage analysis pins down the ceiling of what test-time refinement alone can achieve. Together, these findings map both the capabilities and the limits of LLMs as autonomous scientific problem-solvers.
title AInstein: Can LLMs Solve Research Problems From Parametric Memory Alone?
topic Artificial Intelligence
url https://arxiv.org/abs/2510.05432