# Plausible but Wrong: A case study on Agentic Failures in Astrophysical Workflows

arXiv: https://arxiv.org/abs/2604.25345 · HTML: https://mlsanity.com/paper/2604.25345/

## Key facts

- Cited by 1 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 20 of its 28 references that are in the map.
- First posted on arXiv 28 Apr 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, 20 of 20)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Toolformer: Language Models Can Teach Themselves to Use Tools](https://mlsanity.com/paper/2302.04761/) (2023)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [TruthfulQA: Measuring How Models Mimic Human Falsehoods](https://mlsanity.com/paper/2109.07958/) (2021)
- [WebGPT: Browser-assisted question-answering with human feedback](https://mlsanity.com/paper/2112.09332/) (2021)
- [A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questions](https://mlsanity.com/paper/2311.05232/) (2023)
- [Survey on Evaluation of LLM-based Agents](https://mlsanity.com/paper/2503.16416/) (2025)
- [ScienceBoard: Evaluating Multimodal Autonomous Agents in Realistic Scientific Workflows](https://mlsanity.com/paper/2505.19897/) (2025)
- [TRAJECT-Bench:A Trajectory-Aware Benchmark for Evaluating Agentic Tool Use](https://mlsanity.com/paper/2510.04550/) (2025)
- [LLM-based Agents Suffer from Hallucinations: A Survey of Taxonomy, Methods, and Directions](https://mlsanity.com/paper/2509.18970/) (2025)
- [Large Language Models Hallucination: A Comprehensive Survey](https://mlsanity.com/paper/2510.06265/) (2025)
- [SciToolAgent: A Knowledge Graph-Driven Scientific Agent for Multi-Tool Integration](https://mlsanity.com/paper/2507.20280/) (2025)
- [From Language to Action: A Review of Large Language Models as Autonomous Agents and Tool Users](https://mlsanity.com/paper/2508.17281/) (2025)
- [The Denario project: Deep knowledge AI agents for scientific discovery](https://mlsanity.com/paper/2510.26887/) (2025)
- [AI Agents Can Already Autonomously Perform Experimental High Energy Physics](https://mlsanity.com/paper/2603.20179/) (2026)
- [Open Source Planning & Control System with Language Agents for Autonomous Scientific Discovery](https://mlsanity.com/paper/2507.07257/) (2025)
- [Agentic Artificial Intelligence for Multistage Physics Experiments at a Large-Scale User Facility Particle Accelerator](https://mlsanity.com/paper/2509.17255/) (2025)

## Built on by (most cited first, 1 of 1)

- [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)

## Abstract (arXiv)

Agentic AI systems are increasingly being integrated into scientific workflows, yet their behavior under realistic conditions remains insufficiently understood. We evaluate CMBAgent across two workflow paradigms and eighteen astrophysical tasks. In the One-Shot setting, access to domain-specific context yields an approximately ~6x performance improvement (0.85 vs. ~0 without context), with the primary failure mode being silent incorrect computation - syntactically valid code that produces plausible but inaccurate results. In the Deep Research setting, the system frequently exhibits silent failures across stress tests, producing physically inconsistent posteriors without self-diagnosis. Overall, performance is strong on well-specified tasks but degrades on problems designed to probe reasoning limits, often without visible error signals. These findings highlight that the most concerning failure mode in agentic scientific workflows is not overt failure, but confident generation of incorrect results. We release our evaluation framework to facilitate systematic reliability analysis of scientific AI agents.