# Why LLMs Aren't Scientists Yet: Lessons from Four Autonomous Research Attempts

arXiv: https://arxiv.org/abs/2601.03315 · HTML: https://mlsanity.com/paper/2601.03315/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- Placed in Web, search, and research agents, one of 16 directions in LLM agents and tool use.
- Builds on 19 of its 33 references that are in the map.
- First posted on arXiv 6 Jan 2026.

## Placed in

- LLM agents and tool use / [Web, search, and research agents](https://mlsanity.com/agents/web-search-and-research-agents/)

## Builds on (most cited first, 19 of 19)

- [Dream to Control: Learning Behaviors by Latent Imagination](https://mlsanity.com/paper/1912.01603/) (2019)
- [AlphaEvolve: A coding agent for scientific and algorithmic discovery](https://mlsanity.com/paper/2506.13131/) (2025)
- [The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search](https://mlsanity.com/paper/2504.08066/) (2025)
- [Accelerating scientific discovery with Co-Scientist](https://mlsanity.com/paper/2502.18864/) (2025)
- [Can LLMs Generate Novel Research Ideas? A Large-Scale Human Study with 100+ NLP Researchers](https://mlsanity.com/paper/2409.04109/) (2024)
- [Chain of Ideas: Revolutionizing Research Via Novel Idea Development with LLM Agents](https://mlsanity.com/paper/2410.13185/) (2024)
- [STORM: Efficient Stochastic Transformer based World Models for Reinforcement Learning](https://mlsanity.com/paper/2310.09615/) (2023)
- [The Ideation-Execution Gap: Execution Outcomes of LLM-Generated versus Human Research Ideas](https://mlsanity.com/paper/2506.20803/) (2025)
- [Early science acceleration experiments with GPT-5](https://mlsanity.com/paper/2511.16072/) (2025)
- [EnvBench: A Benchmark for Automated Environment Setup](https://mlsanity.com/paper/2503.14443/) (2025)
- [GenRL: Multimodal-foundation world models for generalization in embodied agents](https://mlsanity.com/paper/2406.18043/) (2024)
- [All That Glitters is Not Novel: Plagiarism in AI Generated Research](https://mlsanity.com/paper/2502.16487/) (2025)
- [SetupBench: Assessing Software Engineering Agents' Ability to Bootstrap Development Environments](https://mlsanity.com/paper/2507.09063/) (2025)
- [Revisiting Epistemic Markers in Confidence Estimation: Can Markers Accurately Reflect Large Language Models' Uncertainty?](https://mlsanity.com/paper/2505.24778/) (2025)
- [Physics Supernova: AI Agent Matches Elite Gold Medalists at IPhO 2025](https://mlsanity.com/paper/2509.01659/) (2025)
- [LLMs Caught in the Crossfire: Malware Requests and Jailbreak Challenges](https://mlsanity.com/paper/2506.10022/) (2025)
- [Learning to Communicate Through Implicit Communication Channels](https://mlsanity.com/paper/2411.01553/) (2024)
- [Expected Return Symmetries](https://mlsanity.com/paper/2502.01711/) (2025)
- [Differentiable Tree Search Network](https://mlsanity.com/paper/2401.11660/) (2024)

## Built on by (most cited first, 13 of 13)

- [Recursive self-improvement of AI research agents](https://mlsanity.com/paper/2609.26457/) (2026)
- [How Do Agents Fail on AutoResearch: End-to-End Diagnostic Evaluation on 100 Real-World Frontier Research Tasks](https://mlsanity.com/paper/2608.14905/) (2026)
- [Self-Improvement of Large Language Models: A Technical Overview and Future Outlook](https://mlsanity.com/paper/2603.25681/) (2026)
- [SciIntegrity-Bench: A Benchmark for Evaluating Academic Integrity in AI Scientist Systems](https://mlsanity.com/paper/2605.10246/) (2026)
- [Distributed Denial of Science: How Indirect Data Poisoning of AI Systems Can Industrialize Scientific Fraud](https://mlsanity.com/paper/2607.10712/) (2026)
- [What is Missing from AI Post-Training AI: An Empirical Analysis](https://mlsanity.com/paper/2608.19072/) (2026)
- [REVERE: Reflective Evolving Research Engineer](https://mlsanity.com/paper/2603.20667/) (2026)
- [One Reflection Is Not Enough: Self-Correcting Autonomous Research via Multi-Hypothesis Failure Attribution](https://mlsanity.com/paper/2606.31478/) (2026)
- [EviGraph: Evidence-Guided Autonomous Research Agents](https://mlsanity.com/paper/2608.04738/) (2026)
- [AutoResearch: Insight In, Hallucination Out](https://mlsanity.com/paper/2608.17906/) (2026)
- [Beneath the Diff: Diagnosing and Mitigating Algorithmic Mode Collapse in Code-Level Autonomous Research Loops](https://mlsanity.com/paper/2609.00077/) (2026)
- [Verify, Don't Trust: Agentic Model Development for Video Discovery Retrieval at Scale](https://mlsanity.com/paper/2609.21257/) (2026)
- [ResearchFlow-Eval: Tracing the Execution Gap in End-to-End Multi-Claim Autoresearch](https://mlsanity.com/paper/2610.18088/) (2026)

## Abstract (arXiv)

We report a case study of four end-to-end attempts to autonomously generate ML research papers using a pipeline of six LLM agents mapped to stages of the scientific workflow. Of these four, three attempts failed during implementation or evaluation. One completed the pipeline and was accepted to Agents4Science 2025, an experimental inaugural venue that required AI systems as first authors, passing both human and multi-AI review. From these attempts, we document six recurring failure modes: bias toward training data defaults, implementation drift under execution pressure, memory and context degradation across long-horizon tasks, overexcitement that declares success despite obvious failures, insufficient domain intelligence, and weak scientific taste in experimental design. We conclude by discussing four design principles for more robust AI-scientist systems, implications for autonomous scientific discovery, and we release all prompts, artifacts, and outputs at https://github.com/Lossfunk/ai-scientist-artefacts-v1